Professional Documents
Culture Documents
Raffaela Giovagnoli
Robert Lowe Editors
The Logic
of Social
Practices
Studies in Applied Philosophy, Epistemology
and Rational Ethics
Volume 52
Editor-in-Chief
Lorenzo Magnani, Department of Humanities, Philosophy Section, University of
Pavia, Pavia, Italy
Editorial Board
Atocha Aliseda
Universidad Nacional Autónoma de México (UNAM), Mexico, Mexico
Giuseppe Longo
CNRS—Ecole Normale Supérieure, Centre Cavailles, Paris, France
Chris Sinha
School of Foreign Languages, Hunan University, Changsha, China
Paul Thagard
University of Waterloo, Waterloo, Canada
John Woods
University of British Columbia, Vancouver, Canada
Studies in Applied Philosophy, Epistemology and Rational Ethics (SAPERE)
publishes new developments and advances in all the fields of philosophy,
epistemology, and ethics, bringing them together with a cluster of scientific
disciplines and technological outcomes: ranging from computer science to life
sciences, from economics, law, and education to engineering, logic, and mathe-
matics, from medicine to physics, human sciences, and politics. The series aims at
covering all the challenging philosophical and ethical themes of contemporary
society, making them appropriately applicable to contemporary theoretical and
practical problems, impasses, controversies, and conflicts. Our scientific and
technological era has offered “new” topics to all areas of philosophy and ethics—for
instance concerning scientific rationality, creativity, human and artificial intelli-
gence, social and folk epistemology, ordinary reasoning, cognitive niches and
cultural evolution, ecological crisis, ecologically situated rationality, consciousness,
freedom and responsibility, human identity and uniqueness, cooperation, altruism,
intersubjectivity and empathy, spirituality, violence. The impact of such topics has
been mainly undermined by contemporary cultural settings, whereas they should
increase the demand of interdisciplinary applied knowledge and fresh and original
understanding. In turn, traditional philosophical and ethical themes have been
profoundly affected and transformed as well: they should be further examined as
embedded and applied within their scientific and technological environments so to
update their received and often old-fashioned disciplinary treatment and appeal.
Applying philosophy individuates therefore a new research commitment for the
21st century, focused on the main problems of recent methodological, logical,
epistemological, and cognitive aspects of modeling activities employed both in
intellectual and scientific discovery, and in technological innovation, including the
computational tools intertwined with such practices, to understand them in a wide
and integrated perspective.
Editors
123
Editors
Raffaela Giovagnoli Robert Lowe
Social Ontology and Social Epistemology Division of Cognition and Communication
Faculty of Philosophy Department of Applied IT
Pontifical Lateran University University of Gothenburg
Rome, Italy Gothenburg, Sweden
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The present volume includes the contributions of the workshop “The Logic of
Social Practices” that took place at the UNILOG2018 World Conference in Vichy,
https://www.uni-log.org/vichy2018 and some other invited ones.
Social practices refer to everyday practices which are routinely performed. They
integrate different types of elements such as bodily and mental activities, material
artifacts, knowledge, emotions and skills. The theme of the book aims to raise the
question about the peculiar “logic” of social practices, and it is broadly conceived
so to allow different responses from various research fields. We can investigate the
relationship/continuation with natural behavior and what are the characteristics of
human social behavior. Moreover, there are important results in the area of AI and
computational where we find different approaches that aim at reproducing the
processes underlying social practices.
From a philosophical perspective, the study of social practices originates from
American Pragmatism (Dewey, James, Mead) and it is relevant to a new conception
of experience. Experience means “practice,” namely the process that characterizes
humans in their continuous relation with their natural and social environment.
According to Dewey, to be human means to be engaged in practices. This thought
is at the center of the so-called practice theory (Schatzki and Reckwitz). It analyzes
important features of the world we inhabit; insofar, it is routinely created using
tools, language and bodily activities. The nature of social practices is at the center
of the debate on collective intentionality (Searle, Tuomela, Bratman, Gilbert).
Collective intentionality represents an important topic in sociology (Durkheim and
Weber), and there are important studies also in the philosophical tradition (Scheler,
Walther, Sellars). Collective intentionality is a mental capacity necessary to create
the social world. There are forms of it also in non-human animals, but the way in
which we create habits, rituals, norms and institutions seems to require a special
research and also possibilities of criticism.
The present volume investigates several aspects of social practices, and it is
intended in a transdisciplinary way. The first part focuses on the relationship
between collective intentionality and social practices. In the ambit of the debate on
collective intentionality, Michael Barber considers the genetic phenomenology
v
vi Preface
decisions or act autonomously. First, we will question which features human beings
and artificial beings share, focusing on the concepts of “person” and “legal per-
sonality,” investigating the philosophical work of thinkers such as Kelsen,
Scarpelli, Radin, Wolff and Nèkàm. The urgency of the (philosophical) matter is
confirmed by the concept of “electronic person” proposed in the motion that the
Committee on Legal Affairs presented for a European Parliament resolution with
recommendations to the Commission on Civil Law Rules on Robotics in order to
establish rules to arrange the relationship between robots and human beings.
The third part concerns the notions of “habit,” “routine” and “ritual” as well as
their interplay. Pawel Kawalec investigates research routines by developing an
evolutionary approach. He aims at identifying some general patterns in social
practices. In particular, he shows three mechanisms underlying cognitive dynamic
of research routines by using the case of microRNA study. In his chapter, Robert
Lowe discusses value-based habitual and goal-directed systems as studied in the
animal and human learning literature. It focuses on the means by which these two
systems might interact in knowledge transfer, particularly as it applies to social
learning. Knowledge is conceived here in terms of types of logic computations as
implemented by neural networks. A discussion of dual-process-type structures in
the brain is provided as well as neural-dynamic implementations thereof and con-
siderations for how a perspective of the brain as carrying out logic computations
might be useful for developing the general cognitive capacities of artificial agents.
Anna Kawalec considers rituals as they are constitutive of a wide range of social
practices. She proposes a conception of ritual that goes beyond traditional patterns
and disciplinary boundaries. An original “folk” conception aims to be functional by
crossing philosophy and anthropology. She examines selected categories and
concepts of philosophy and anthropology to substantiate her case: (i) the etymo-
logical meanings in Sanskrit, (ii) the Hebrew origins, (iii) the ethnographic facts (on
the basis of Alfred Gell’s research on Muria’s “God playing” and some selected
theses of cultural and anthropological theory), (iv) Jerzy Grotowski’s idea of
techniques of the performer’s breath and (v) some points from Karol Wojtyła’s
philosophical integrative view of the human person and agency. Habits and rituals
represent a relevant field to investigate human behavior also in comparison with
non-human animals and nature in general. Raffaela Giovagnoli moves from very
interesting perspectives that cross philosophy and neuroscience to establish a
continuity between individual and social habits and to clearly isolate what makes
human rituals peculiar in the field of social ontology. Therefore, she introduces a
general and brief discussion on rituals in different disciplines before focusing on the
actuality of the Aristotelian notion of habit in relationship with some studies in
neuroscience that can be extended to the social context. The perspective of
we-intentionality introduced by John Searle embeds the dimensions of habits that
ground cooperation and is very useful to describe the peculiarity of human rituals.
ix
x Contents
Michael Barber
Abstract There has been an extensive debate about we-intentionality, with some
positions demonstrating how it is derived from individual intentionality or others
arguing that it precedes and takes priority over individual intentionality. Recent con-
tributions have come to recognize that the theoretical discussions of we-intentionality
depend upon pre-theoretical, pre-analytic, and commonsense experiences of joint
intentionality, the description of which might avoid some of the polarizations that
the theoretical discussions cannot escape. One needs to consider the spectrum of posi-
tions from the more individualistically inclined treatments (Bratman, Tuomela) to
those attempting to establish the irreducibility of we-intentionality (Searle, Gilbert) to
those de-emphasizing individual intentionality (Baier) or striving for a more refined
balance (Schmid). In the end, Alfred Schutz’s genetic phenomenology, capturing
pretheoretical, commonsense experience, can show how we-intentionality is a prim-
itive experience in which individual consciousness is at the same time historically
and biographically shaped. Schutz’s account holds in tension the individual and
social dimensions of experience and avoids reducing we-intentionality to individual
intentionality.
1 Introduction
Within analytic philosophy, there has been an extensive debate about the character
of we-intentionality, between positions in which we-intentionality is derivative from
individual intentionality or in which the former takes precedence over the latter.
Several of the discussants in this debate have come to recognize the importance of
pre-theoretical, pre-analytic, commonsense experiences of joint intentionality whose
M. Barber (B)
Saint Louis University, St. Louis, MO 63108, USA
e-mail: michael.barber@slu.edu
© Springer Nature Switzerland AG 2020 3
R. Giovagnoli and R. Lowe (eds.), The Logic of Social Practices,
Studies in Applied Philosophy, Epistemology and Rational Ethics 52,
https://doi.org/10.1007/978-3-030-37305-4_1
4 M. Barber
consideration might enable one to escape some of the dilemmas and polarizations
engendered in theoretical discussions. Genetic phenomenology,1 particularly that
developed by Alfred Schutz, can show how we-intentionality is experienced as a
primitive experience in which individual consciousness, at the same time histori-
cally and biographically shaped, is developed. This genetic account holds in tension
the individual and social dimensions of human experience as well as the possibility
of a we-intentionality that is not reducible to individual intentionality. The positions
on joint intentionality vary from positions that stress the individual origins of we-
intentionality (such as Michael Bratman’s and Raimo Tuomela’s) to those that seek,
without complete success, to establish the irreducibility of we-intentionality (John
Searle and Margaret Gilbert). By contrast, other have sought to de-emphasize indi-
vidual intentionality (Annette Baier) or achieve a kind of balance (Hans Bernhard
Schmid) that would be more fully accounted for through a genetic phenomenological
approach. First, I will consider those inclined toward the individual origins (Bratman
and Tuomela) as well as those who strive to uphold we-intentionality, even though
they still seem entrapped within a kind of individualism (Searle and Gilbert). Then I
will turn to those who stress the independence and even priority of we-intentionality
(Baier, Schmid).
1 “Genetic phenomenology” can have several senses. In this paper and in the work of Alfred Schutz,
the “we relationship,” is something experienced, first of all, in the everyday life-world that is the
often-forgotten basis that higher-level theories arise out of and often forget and that we all inhabit
before we undertake theorizing. In addition, it is within that everyday life we-relationship that the
child begins learning about how to live and maneuver in the life-world. In each sense, that of the
life-world and that of the child’s development, one is starting with something already established,
the theoretical sphere or the experience of adults, and reconstructing experiences that genetically
preceded those higher-level activities and that gave them birth.
Joint and Individual Intentionality: A Genetic … 5
the I-mode, thinking “I intend to contribute to the group’s goal.” [2]. In the we-mode
of joint attention, one thinks of oneself as doing one’s part in tandem with others in
the group, one feels entitled to expect of others that they do their part, one expects
to be exposed to critique or to criticize others for any arbitrary withdrawals from the
group action, one believes that these commitments are public and mutually known,
one monitors others as they monitor oneself, and those engaged in pursuing in the
we-mode a collective goal express their collective acceptance of the group goal as
“Our goal is p” [2] or “We will achieve p.” [2]. All these interactional entwinements
contrast with group dynamics when group members typify themselves as “private
persons” with only aggregated private commitments [2].
Given the depth of involvement and the intersubjective intricacy that we-mode
joint action calls for, in contrast to I-mode, one can see why Tuomela can claim
that his notion of “we” is not reducible to the conjunction “you and I” [2]2 and
why his theory is “conceptually rather anti-individualistic, even if ontologically it
is, if not individualistic, at least interrelationistic and eschews collective agents and
actions in a literal ontological sense” [2].3 Further evidence of his less individualistic
orientation is to be found in the fact for those in a we-mode the collective action
dilemma that one might find in highly individualistic prisoners’ dilemma scenarios
do not exist insofar as group members do not “act strategically with respect to each
other but do as the group reason specifies,” [2] as opposed to what might occur in the
private or I-mode point of view. In addition, Tuomela criticizes Michael Bratman’s
analysis of shared intentions in which an “I” intends that a “we” engage in a joint
action and a “you” intends that “we” engage in this action and in which our subplans
happen to mesh and we have common knowledge of all these factors [3]. In contrast
to Bratman’s explanation, Tuomela presents an account of joint action in which I
intend to and would be committed to do my part just as you are, in which both
2 Another way in which Tuomela argues that a we-intention differs from and is irreducible to an
individual intention is by distinguishing an “aim-intention” in contrast to an “action intention.”
He states, “An aim-intention can be satisfied without the aim-intending agent alone satisfying the
intention, whereas in the case of an action intention the agent must believe, if rational at all, that
he can (with some probability) satisfy the intention by means of his own actions.” See Tuomela
[1]. Aim-intentions are clearly not reducible to the action intentions of individuals and the “we”
is not reducible to “you and I,” nevertheless the way aim intentions are developed depends upon
individuals taking account of, considering, or estimating whether others will be willing to an adopt
an end and “do their part” to realize it [1].
3 Tuomela [1, 2]: One ought not to think that there can be no group, joint action if group members
do not share a we-intention insofar as Tuomela allows for joint intention in a we-mode sense and
an I-mode sense. The We-mode includes a sense of “for-groupness” and acting for a group reason,
whereas the I-mode requires only private reasons for cooperation; and in the we-mode the group
has authority whereas in the I-mode each person has full authority over whatever he or she is doing;
in we-mode, the group members are committed to each other as opposed to the “I-mode” in which
the participant is only committed to herself; in the We-mode actors are willing to do all parts if
others do not agree but the I-mode would not do the same and instead weighs actions in terms of
individual costs or private reasons. Tuomela insists that his analysis is conceptually non-reductive
since all these intertwined conditions must be in place as opposed to having individual intentions
merely aggregated with each other, and he admits that his account is “ontically individualistic or,
better, interrelational” [1].
6 M. Barber
of us are aware of each other’s willingness to do our parts, and in which these
mutually shared beliefs serve in part as reasons for my intending to do my part of our
common action [2]. In Tuomela’s opinion, for Bratman the agents have their personal
intentions on the basis of their private reason and there is, in effect, no group reason
(in accord with which each partner executes his particular part of the joint action
with another). Bratman’s account represents simply an I-mode willingness to act,
but without the kind of cooperation that a joint action would require, such that if
the other person did something wrong one might step in (“do one’s part”) to ensure
that the action intended, to which both parties are committed, is realized [2]. Further,
Tuomela allows that we-mode shared intentionality can be found among not yet fully
developed individuals, such as infants, or it can be produced in conformist, cultural
transmission among individuals who are not clearly individuated, differentiated, and
mature, as if autonomous independent agents, fully in possession of themselves, are
not necessarily a pre-condition of joint action [2].
However, the fact that Tuomela’s standpoint includes a much more expansive
we-mode of group participation in contrast to Bratman’s I-mode does not imply
that he is anti-individualist. His notion of we-mode joint intentionality seems to
presuppose autonomous individuals who voluntary choose to enter into a we-mode
joint action, and whether or not one will perform one’s part as part of a joint action
depends on how such individuals typify themselves, that is, what role they take up
with regard to a group, for instance, as a group member or as a private person [2].
While legitimately resisting any idea of that a group acts as an ontological agent [2],
Tuomela, it would seem, places a priority on individuals from whom the level of
joint attention is derivative. Hence he asserts that the “we mode is in the minds and
actions of individuals,” [2] that “groups are agents and persons only in a metaphorical
sense,” [2] and that “Much of what is said in this book prima facie treats groups as
real but nonagentive social systems supervenient on the members functioning qua
group members.” [2].
Counter to the individualistic leanings of analytic accounts of shared intention-
ality, such as Tuomela’s and Bratman’s, Annette Baier [4], while acknowledging
that voluntary associations do in fact depend upon “pre-planned intentional ‘joining’
(joint activities) or some intentional ‘sharing’ (shared activities), or by a ‘unification’
of separate wills,” argues that we engage in many activities done with others before
doing them alone. It is precisely this fact that Baier charges that analytic philosophers
of action overlook insofar as they operate with the following assumptions:
One such assumption is that the analysis of singular-person action must precede that of
plural-person action. Another is that intentional actions that initiate something should be
analysed before intentional continuings of what is already started. The third is that “prior
intentions” should be analysed before intention in action [4].
John Searle, as Baier [4] acknowledges, seeks to do justice to the idea of plural
subjectivities, and posits a collective intentional behavior that is not the summation
of individual intentional behavior, in which one’s intention is entirely independent
of the intentions and behavior of others. Rather collective intentionality, as might be
found in an orchestra, corps de ballet, or football team, is based on the we-intention
Joint and Individual Intentionality: A Genetic … 7
expressed as “We are doing act A,” from which individual intentionality, expressed
by “I am doing act A,” is derivative [5]. Searle posits we-intentions as a primitive
form of intentionality and contends that analyses that reduce such intentionality to
individual intentionality (including mutual beliefs about each other’s intentionality)
do not provide sufficient conditions for cooperation [5]. Just as Tuomela opposes any
we-subject, Searle insists that “society consists entirely of individuals, there cannot
be a group mind or group consciousness. All consciousness is in individual minds,
in individual brains” [5]. He also allows that one could be mistaken about what
is actually happening—even to the point that two parties could share a collective
intention if they were merely brains in a vats. Such a collective intention could be
expressed as “We intend that we perform act A” [5] and such an intention can exist
in the mind of each individual agent who acts as part of this collective and whose
individual acts are part of the collective act [5].
Hans Bernhard Schmid criticizes Searle for restricting collective intentionality to
what is in individual minds, and he imagines an evil scientist who might separate two
friends contemplating Matisse’s “Dance,” place their brains in vats on different floors
of the museum in which they were located, and then provide appropriate input so
that both think that they are contemplating Matisse’s “Dance” together. But Schmid
objects,
Intentionality does not become shared intentionality just because completely independently
of each other, two brains just happen to have appropriately ‘matching’ illusions. If shared
intentionality is not a matter of what goes on inside an individual head, it is not a matter of
what goes on inside different heads either… it is not enough to check only what is in the
minds of two individuals … sharedness is a matter of the relations between minds [6].
On the other hand, one can detect in Gilbert’s work a continued kind of emphasis
on individualism, as Baier [4] charges. Gilbert states that joint commitment typically
consists in two people “in face-to-face contact, mutually expressing their readiness
to be jointly committed, in conditions of common knowledge” [7].4 In addition, for
Gilbert those “understandings, expressive actions, and common knowledge” [7] that
replace any “group mind,” [7] play a role in upholding the importance of individuals
in plural action, in particular, the obligations and entitlements partners experience in
relationship to each other in a shared intention. Within a joint commitment each part-
ner is entitled to the corresponding partner’s obligated performance and is licensed
to rebuke that partner if he or she violates such an obligation [7].5 That Gilbert means
these mutual obligations and entitlements between those jointly committed to protect
the individualism of the parties, who are each held to be individually accountable to
each other, appears in the fact that she appeals to these obligations and entitlements
in countering what she takes to be the collectivism of Searle and Baier. Countering
Searle’s view that “We intends” are in the minds of individual human beings with her
own view of how expressions of readiness yield a joint commitment, she proceeds to
emphasize repeatedly that Searle needs to explain his idea of “We intends” through
the mutual obligations and entitlements that are associated with shared intention and
action and that by the partners’ mutual answerability would help preserve the indi-
viduality of each individual [7]. Likewise, in response to Baier’s objection to her and
others’ individualistic bias, she faults Baier for not explaining how doing things with
others, which is more basic than doing things alone, cashes out in terms of the mutual
“accountability those acting together perceive themselves to have to one another” [7].
Baier, in turn, substantiates her accusation about Gilbert’s individualism by citing
her claim in On Social Facts that plural subjects “pool their wills” together [4, 17].
Having seen repeatedly Annette Baier’s critique of those who assign to we-
intentionality a derivative status, it is not necessary to reiterate her arguments above
that highlight the prioritization of individual intentionality. Instead, as an example
of another position stressing we-intentionality, one can consider Bernhard Schmid’s
effort to defend the priority of we-intentionality through his account of plural action.
Schmid begins by distinguishing intentional autonomy, according to which indi-
viduals perform their own actions, from intentional individualism or motivational
autarky, according to which each individual’s behavior “has to bottom out in his or
her own volitions or pro-attitudes (rather than in some other individual’s volitions or
pro-attitudes)” [6]. Schmid [6] argues that plural actions occur when several persons
share one goal (not just a similar goal), and he suggests that that intentional autonomy
does not preclude intentional hertararky, a kind of other-regarding motivation that
does not bottom out in one’s own volitions (or motivations) and that does not reduce
oneself to falling under the remote control of another or to acting as an intentional
zombie. Schmid provides examples:
What is at stake here are simple patterns such as the following: holding open a door for a
stranger, spontaneously helping a stranger to lift a baby carriage into a train, or moving aside
a little on the park bench so another person can find a seat too… These are social actions,
because they require cooperation, and they are plural actions exactly insofar as the helper’s
goal is the same as the individual’s who is being helped [6].
At this point, favoring a “teamwork” model [6] and concurring with Baier that we
have been brainwashed by Cartesian individualism [4, 6], Schmid rejects Bratman’s
individualism as well as what he calls Searle’s subjective individualism that would
permit joint intentionality between brains, placed in vats, which are not really in
relationship with each other [6], favoring a form of collective intentionality that is
irreducible to individual intentionality and yet involves real relationships between
individuals. Contrary to the positions of Bratman and Tuomela then, Schmid makes
the case that collective intentionality, or shared intentionality, actually takes prece-
dence over any individual intending, such as one’s intending to do one’s part, any
decision to share intentionality, any agreement to share intentionality, or any under-
taking of the kinds of entitlements or obligations that Gilbert discusses. For instance,
if one, like Gilbert, posits tacit agreement between two individuals at the basis of
shared intentionality, Schmid points out that a pre-condition for entering into such an
agreement is that a certain kind of shared intentionality already must be in place so
that the two parties can come together to agree. Agreeing itself is something people
have to do together [6]. For Schmid, it is possible to have shared intentional activities
10 M. Barber
without any obligations and entitlements being operative and hence shared intention-
ality is not itself socially normative, though it inevitably leads to socially normative
consequences. Hence, merely habitual social practices, like customs, already involve
shared intentionality, but they usually issue in social norms. In Schmid’s opinion,
having a collective intention, then, provides a reason to develop an appropriate per-
sonal intention, such as to do one’s part in a common endeavor [6]. Schmid provides
numerous examples of people sharing intentionality in which one lacks a sense that
one participates as a distinct individual. One instance appears when two colleagues
collectively intend to go to a cafeteria and one of them becomes so absorbed in the
conversation that she forgets that they are even going to the cafeteria or that she had
agreed to go to the cafeteria in the first place. The colleagues share an intentionality
in which at least one party’s individual intentionality has dropped from sight [6].
Or at a public assembly the president of an organization might express “our” (the
company’s) grief over the death of one of its officials, even though those attending
and being interviewed afterward, might admit to feeling little of the grief that the
president spoke of. There is a shared intentionality even though some members do
not experience themselves as individual participants [6]. To head off any concern that
such shared intentionality might imply a collective super-agent, Schmid urges read-
ers not even to ask the question who the subject is who has the collective intention;
the question itself is misplaced [6].
initiate, though Gilbert accuses Baier of neglecting to clarify the individual that her
system of entitlements and obligations reveals. Finally, Schmid posits shared inten-
tionality as the fundamental from which individual intentionalities emerge, gradually
including the kinds of normative social relationships that Gilbert accentuates, and
he explains how intentional autonomy need not preclude intentional heterarky. On
a spectrum one might plot Bratman and Tuomela at the individualist pole, with
Gilbert and Searle in intermediate position, and Baier and Schmid at the more col-
lectivist pole. What becomes interesting, however, is the ways in which many of the
authors considered seek to return to a level of experience beneath that of their the-
orizing, to seek to recover how actors experience their individual intentionality and
shared intentionality, an example of which occurs in Searle’s above mentioned idea
of “pre-intentional background conditions.” For instance, Tuomela, at the end of his
essay “We-Intentions Revisited” responds to criticisms that his argument is circular
because he seeks to define collective intention as each “doing his (or her) part” and
yet the very idea of doing one’s part presupposes collective intention. Tuomela then
attempts to describe how pre-analytic, ordinary, common-sense agents might think
about joint intention even though they might not have in view at all the notion of
collective intention that is central to the theoretician’s point of view, that is described
above in terms of an individual actor we-intending to perform an action, and that
involves an intricate intertwined taking account of other’s willingness to do their
part, etc. Such a common-sense actor “intends to perform X with the others in part
because (of his belief that) the others intend to perform their parts of X and intend
to perform X with the others” [1]. It should be noted that this description of the
common-sense actor’s point of view seems more individualistic than ever since one
seems to take even less account of beliefs about others’ beliefs, and, at the same time
it consists a kind of individualistic calculation that one’s own action is conditional
upon the cooperation of others. Nevertheless, this effort to reconstruct how common
sense actors experience joint intentions converges, as I hope to show, with a kind of
genetic phenomenology that Alfred Schutz provides and that synthesizes both the
collective and individualistic dimensions of joint intentionality.
Similarly, John Searle, at the end of “Collective Intentions and Actions,” after
providing his own account of collective action, turns to discuss a pervasive “sense
of community” [5] that collective intentionality itself presupposes and that includes
the kinds of background skills needed, for instance, to play football together or per-
form in an orchestra together. Searle argues that such background features are not
easy to characterize, that old-time philosophers tried to get at them by describing
us as social or political animals, and that they require a “pre-intentional sense of
‘the other’ as an actual or potential agent like oneself in cooperative activities” [5].6
6 Itshould be noted that when Searle speaks of a “pre-intentional” sense of the other, he is taking
“intentional” in the typical way that analytic philosophy uses that term, namely as describing
an aiming at a goal toward which action will be directed and as implying a certain amount of
deliberateness. For Schutz and the phenomenological tradition, “intention” refers to any conscious
act, such as desiring, thinking, memory, judging, etc. that is aimed at an object (and hence one is
desirous of, thinks of, has a memory of, or judges about), and often one is unaware of such intentions
until one reflects upon them.
12 M. Barber
About such background features, Searle posits that one just presupposes them and
that one does not have any special belief about them being there, in the same way
that one supposes without specifically believing that the ground beneath me is solid
and that others are conscious agents. For Searle we cannot ultimately explain soci-
ety through conversation or other collective behaviors since even they presuppose
a form of society for them to function at all. In my view, Searle’s forays into this
question precisely converge with the life-world that Schutz and Luckmann describe
as that “province of meaning which the wide-awake and normal adult simply takes
for granted in the attitude of common sense” [9]. It is “unquestionable” [9] and “un-
problematic until further notice” [9]. In addition, Schmid seeks to cast doubt on the
idea that collective intentional activity is based in some commitment or dedication
to a cause on the part of “participating individual,” [6] and he seeks to recover “intu-
itive, pre-theoretic understanding of what it means to share an intention” [6] “without
implying any appropriate individual contribution or contributive intentions, or any
individual we-intending” [9]. As we have seen, Schmid [6] offers examples of shared
intentionality such as when one of the individuals has completely forgotten about
the intentionality with which the joint action was initiated or when a group shares an
emotion despite the fact that some participants do not feel it—both of which illustrate
“non-participatory uses of we” [6]. Given these examples and given Schmid’s obser-
vation that “social norms arise out of merely habitual social practices” [6] (in which
one shares intentionality before finding oneself obligated or entitled as an individ-
ual to certain actions), Schmid contends that “the analysis of shared intentionality
cannot be based on an analysis of what individuals personally intend when sharing
an intention” [6]. This pointing toward non-participatory shared intentionality, as
Schmid describes it, can indicate, I hope to demonstrate, the kind of experiences
that Schutz’s account of the life-world brings into focus. Finally, the author whose
thought most approaches a genetic phenomenology like Schutz’s is Annette Baier
in “Doing Things with Others: The Mental Commons.” Baier’s ultimate premise is
that shared intentions lie at the root of mutually referring individual intentions and
that individual action is itself learned as a departure from common action. Critical
of Jon Elster’s individualism, she posits that social norms become mysterious if they
are supposed to be “evolved either by chance or by rational agreement between fully
fledged self-interested individualists” [4]. Allowing a role for individual intentions in
voluntary associations in which pre-planned intentional action might be important,
Baier wonders why “first person singular should be deemed more fundamental than
first person plural” [4]. Instead, she focuses on unplanned actions rather than planned
ones, on inevitably intersubjective institutional actions, on unreflective customs that
structure unpremeditated acts, on activities that two people find themselves doing
together (such meeting each other’s gaze, glaring at each other, traveling alongside
another), on non-voluntary activities on which the voluntary are parasitic, and on the
intentional continuation of activities that were not intentionally intended or planned
beforehand. Instead of considering actions that one does on one’s own and might
admit others to partake of (e.g. such as sharing one’s bank-account), she concentrates
on actions one cannot do alone or learn to do alone such as war-making, wrestling,
Joint and Individual Intentionality: A Genetic … 13
Schutz, as a philosopher, defines the we-relationship (in a way that the common-
sense actor would never think of doing) as a matter of each partner to a relationship
attending in simultaneity to the ongoing processes of each other in a face-to-face
relationship, in which we share the same space, bodily presence to each other, and
time—or, as Schutz puts it, a “vivid present” [11]. Schutz demonstrates how such
a we-relationship is enacted in the simplest conversation in which one follows a
partner’s building up the thought being communicated, through a process that unfurls
with each participant simultaneously coordinating with each other, step by step, word
by word. Such a we-relationship, based on a sharing of space, time, and the unfolding
process permits both parties to say “We experienced this occurrence together” [11]
14 M. Barber
or to comment that “we grow older together” [11]. While one lives immersed in the
we-relationship with the other, one is absorbed with the development of the other’s
conscious processes in the fullness of their concrete content in the vivid presence
one shares, and to bring that relationship into reflective focus, the partners must stop
focusing on the other and step out of the we-relationship [12].
The child, despite lacking much of a developed self or of consciousness, clarity,
or determinateness about what is going on, engages in a we-relationship with adults
who conduct themselves as if they were in a we-relationship with the child. Only
in this context, in this “reciprocal motivational context,” [9] does the child slowly
come to understand goals, means, attitudes, identity, and the group’s natural world
view, and to develop his or her own personal identity [9]. Furthermore, “the concrete
reality of the we-relationship is the social basis for the individual learning a language
as a system of meanings referring to ‘any reality whatever’” [9]. Accentuating the
importance of such a shared relationship, Schutz agrees with Max Scheler’s objection
to the supposition of the analogy argument for the existence of the Oher that the first
thing given to us is our own self. Instead Schutz concurs with Scheler that the “We”
is naively presupposed and given to each of us as prior to the I. In fact, for Scheler
psychological accounts of children and “primitives” indicate that they live in others’
experiences before living in their own in such a way that one cannot differentiate
one’s experiences from those of another. Schutz accepts that we live in others’ life
instead of our own but only insofar we are unreflectively directing our thoughts and
acts toward the other in the “vivid simultaneity of the We” [13].7 However, when one
turns in reflection to one’s experience and breaks the spell of the We-relationship, it
becomes clear that there is no such thing as an experience given to me that does not
indicate to which individual stream of consciousness the experience belongs. Schutz
agrees that children and “primitives” exhibit the immersion in an experience in which
our individual experiences may appear indistinguishable—as we all do while living in
the we-relationship—but this is also in part because children and “primitives” acquire
later the capacity for reflection, in which I, like them eventually, once outside the
bounds of the we-relationship, recognize my stream of consciousness as “through
and through the stream of my experiences” [13].
Having presented the workings of the we-relationship, Schutz and Luckmann raise
the question of how it is “constituted,” [9] and this technical, phenomenological term
refers, not to a kind of idealistic creation of something that really exists only in one’s
mind, but rather to the fact that, when one experiences an object as something of a
finished product standing before oneself, say a house, it is always possible to inquire
back into the many types and levels of experiences that are involved in experiencing
the object; that would go into the building up of final meaning of the object, as a house,
for instance; that one may have forgotten or never even focused on; or that one might
reconstruct in attempt to give a systematic account of how the meaning of the object
7 Ofcourse, the use of the term “primitives” to describe cultures different from one’s own or not as
technologically advanced as one’s own is regrettable. The negative connotations of that term may
lead one to ignore the psychological or spiritual dimensions of such cultures that may be richer than
those in one’s own culture.
Joint and Individual Intentionality: A Genetic … 15
8 Therewould be several indications that the other is “like” me, and the indications of the likeness
and the likeness itself are not a matter of inference, but passive syntheses, by which I immediate
assimilate the other to myself, as capable of sharing a we-relationship.
16 M. Barber
But the shared intentionality (in the phenomenological sense) in the we-
relationship is not only a matter of the partners mutually immersed in each other’s
unfolding conscious processes and the mutual coordination between them, but it is
also possible for them to jointly attend to an object in their environs. For instance,
Schutz comments on how two partners can “grow older together” by simultaneously
experiencing together in a we-relationship a bird in flight. Schutz the phenomenolo-
gist reflectively unpacks the constitutive underpinnings of this experience that either
of us might report in terms “we have seen a bird in flight.” He observes, for instance,
that “Perhaps, while I was following the bird’s flight I noticed out of the corner of
my eye that your head was moving in the same direction as mine” [12]. And, the
statement about us seeing the bird would depend on many other intentional linkages
about which neither of us may be explicitly conscious (e.g. recognizing that you
are an Other like me, that you are not visually handicapped, that your body overall
indicates that you are wide-awake and alert instead of sleep walking—all of which I
may be vaguely aware at the moment when we see the bird’s flight together). But in
spelling out the underlying intentional anticipations (and their confirmation insofar
as they continue to be borne out), we are only giving a constitutive analysis of the
experiences, of whose complexity both parties may not even be conscious and in
which one may have no idea of the individuality of one’s intentionality as distinct
from the other’s. The experience that Schutz analyzes consists in the jointly intended
perceptual act in which, as would be reported by both the participants, “we together
see a bird a bird in flight.”
In addition to a shared perceptual intentionality with the same perceptual object
that we are perceiving together, a “we” can jointly intend a project in which we are
so immersed that we are not mindful of our individual activity and our individual
contribution, so taken up are with each other and what we are doing together. Schutz
points to a series of activities in which this is the case: playing baseball or tennis,
fencing, dancing together, making love together, or making music together [15].
He articulates how a “we” intend together the last-mentioned project insofar as
each of the co-performers of a group undertake the project of producing a musical
presentation, in which each
has, therefore, to take into account what the other has to execute in simultaneity. He has not
only to interpret his own part which as such remains necessarily fragmentary, but he has also
to anticipate the other player’s interpretation of his—the Other’s part and, even more, the
Other’s anticipations of his own execution. Either’s freedom of interpreting the composer’s
thought is restrained by the freedom granted to the Other. Either has to foresee by listening to
the Other, by protentions and anticipations, any turn the Other’s interpretation may take and
has to be prepared at any time to be leader or follower. Both share not only the inner durée in
which the content of the music played actualizes itself; each, simultaneously, shares in vivid
present the Other’s stream of consciousness in immediacy. This is possible because making
music together occurs in a true face-to-face relationship—inasmuch as the participants are
sharing not only a section of time but also a sector of space. The Other’s facial expressions,
his gestures in handling his instrument, in short all the activities of performing, gear into
the outer world and can be grasped by the partner in immediacy. Even if performed without
communicative intent, these activities are interpreted by him as indications of what the Other
is going to do and therefore as suggestions or even commands for his own behavior. Any
Joint and Individual Intentionality: A Genetic … 17
chamber musician knows how disturbing an arrangement that prevents the coperformers
from seeing each other can be [15].
relationships of joint intentionality. One cannot dismiss the individuality that comes
to the fore when one decides to enter into the voluntary associations that Tuomela
focuses on or that appears in the notion of intentional autonomy that Schmid refuses
to let go of, despite Schmid’s proclivity for “non-participatory” we-experiences that
converge with Schutz’s position on “we-relationships.” One ought not overlook the
presence of individuality even within Schutz’s depiction of the we-relationship,
despite the fact that such a relationship in a family subsumes within it children
whose self-identity is scarcely developed and whose parents in effect “force” [9]
or “impose” [9] the reciprocity of the we-relationship on them. In addition, one
cannot dispense with individuality despite the fact that the implementation of the
we-relationship among adults tends to be so focused on the other with whom one
shares a vivid present and on the other’s conscious processes coordinated with one’s
own, that the individual self hardly becomes visible at all.
But in the case of Schutz, whose own thought derives from Husserl’s, including the
latter’s emphasis on the transcendental ego individualized by its history as an ongoing
stream of consciousness, one might expect that he would not neglect the importance
of individuality. Indeed, as Schutz himself observes, my subjective experience and
the other’s are such that in my understanding of the other’s experience I can never
understand it the way the other does. I might have been able to do so if I had lived
through all her conscious states and intentional acts, down to the smallest details,
impressions, and horizonal areas of protention and retention. But I would also have to
remember all her experiences and to have lived through them in the order and with the
intensity that she did, paying attention to each of them as she did. As a consequence,
such “‘Intended meaning’ is therefore essentially subjective and is in principle con-
fined to the self-interpretation of the person who lives through the experience to be
interpreted. Constituted as it is within the unique stream of consciousness of each
individual, it is essential inaccessible to every other individual” [12]. Although any
or all our interpreting concepts or thoughts rely on the socially transmitted and inher-
ited typifications bestowed upon us within the various we-relationships to which we
have belonged from childhood on, they are also inflected and informed by the unique
history of the individual relying on them. Schutz insists, though, that this does not
deny that anyone can understand another person’s experience, but that, rather, the
meaning I give to another’s experience cannot be exactly the same as the other’s [12].
Schutz and Luckmann recognize that the subjective system of interpretational and
motivations relevances, including plans, typifications, models of explication, and so
on, have a “prior history” [9] that one brings into every we-relationship. They reit-
erate over and over again that same events and objects never have exactly the same
meanings for one partner as they do for another because of the unique biographi-
cal situation that each brings into any we-relationship [9]. Indeed, one can imagine
that even in one’s initial we-relationship, in the pregnancy that Baier references,
should the mother’s pregnancy be tumultuous, the child might emerge to a degree
as more physically and psychologically intense than a child emerging from a more
tranquil pregnancy. And this very intensity, bodily impressed on the child in the
we-relationship of pregnancy, will affect the next we-relationships into which the
child enters, e.g. in breast feeding or in encounters with one’s father or grandparents.
Joint and Individual Intentionality: A Genetic … 19
In effect, one’s individual biography and identity and one’s inescapable social con-
nectedness and formedness are inextricably intertwined and one’s individuality and
intersubjective relatedness are co-originary from the start. As Schutz suggests else-
where, our biographically situation is an imposed relevance, it is there as given and
we cannot redo it, although we can come to terms with in, given the set of relevances
and typifications we have at hand, and of course that system of “intrinsic relevances”
by which we come to terms with imposed relevances, are themselves the product of
our social and individual history that has preceded the present moment [16].
Schutz, in effect, has reconstructed a genetic account of joint intentionality—
something sought inchoately by the philosophical interlocutors mentioned above—
though, Schutz does so by understanding intentionality phenomenologically, that is,
as including the wide spectrum of acts or orientation toward objects, persons, and
events that are often directed to their objects often without being consciously the-
matized. He shows the foundational character of the we-relationship, for all other
categories of human existence, and the trajectory of one’s biography, which is co-
originary with one’s earliest we-relationships. Schutz’s account then would imply
that individuality and social pertinence are present in any social interactions, though
one comes to the fore over the other in different contexts and in the considera-
tion of different problems. For example, when one is self-consciously deliberating
about whether to engage in a joint action with others, about whether her doing her
part will be matched by them doing their part, and about whether there is sufficient
mutual knowledge for the joint action’s successful execution—all of which precede
embarkation upon a joint action in Tuomela’s view—it is clear that one’s individual
intentionality emerges into prominence and the joint intention appears to be deriva-
tive. Schutz himself has developed an account of how individuals intend a project by
imagining in future perfect tense the project completed, choosing it, and setting in
motion the sub-acts leading to its realization—and such an account could be extended
to explain joint intentions. However, even in those moments when one deliberates
before a joint action, one’s understanding of and willingness to embark with others
on such an action depends on one’s massive past experience of we-relationships,
however much this experience might lie on the horizons of one’s focus while one
deliberates. Likewise, Baier’s explanation of shared intentions, that Gilbert thinks
overlooks individual intentionality and that consist in such activities as pregnancy,
breast feeding, playing a game, or painting a house, simply does not attend to the indi-
viduating stream of consciousness that belongs to each member of a we-relationship
and that is co-originary with that member’s earliest experience of we-relationships.
6 Conclusion
In brief, Schutz recovers the originary, commonsense social bonds of the we-
relationship through a genetic analysis, toward which Searle, Baier, and even
20 M. Barber
References
5. Searle, J.: Collective intentions and actions. In: Cohen, P.R., Morgan, J., Pollack, M.E. (ed.)
Intentions in Communication, pp. 401–415. The MIT Press, Cambridge, MA and London,
England (1990)
6. Schmid, H.B.: Plural Action: Essays in Philosophy and Social Science. Springer, Dordrecht
(2009)
7. Gilbert, M.: Sociality and Group Responsibility: New Essays in Plural Subject Theory. Rowman
and Littlefield Publishers Inc., Lanham (2000)
8. Husserl, E.: Ideas pertaining to a pure phenomenology and to a phenomenological philosophy,
vol. 1. General Introduction to a Pure Phenomenology (trans: Kersten, F.). Martinus Nijhoff,
The Hague (1983)
9. Schutz, A., Luckmann, T.: The Structures of the Life-World, vol. 1 (trans: Zaner, R.M., Tristram
Engelhardt, H.). Northwestern University Press, Evanston (1973)
10. Schutz, A.: The problem of transcendental intersubjectivity in Husserl. In: Schutz, I. (ed.)
Collected Papers 3: Studies in Phenomenological Philosophy, pp. 51–91. Martinus Nijhoff,
The Hague (1966)
11. Schutz, A.: On multiple realities. In: Natanson, M. (ed.) Collected Papers 1: The Problem of
Social Reality, pp. 207–259. Martinus Nijhoff, The Hague (1962)
12. Schutz, A.: The Phenomenology of the Social World (trans Walsh, G., Lehnert, F.). Northwest-
ern University Press, Evanston (1967)
13. Schutz, A.: Scheler’s theory of intersubjectivity and the general thesis of the alter ego. In:
Natanson, M. (ed.) Collected Papers 1: The Problem of Social Reality, pp. 150–179. Martinus
Nijhoff, The Hague (1962)
14. Cooley, C.H.: Human Nature and the Social Order, rev. edn. Charles Scribner’s Sons, New
York (1922)
15. Schutz, A.: Making music together: a study in social relationship. In: Brodersen, A. (ed.)
Collected Papers 2: Studies in Social Theory, pp. 159–178. Martinus Nijhoff, The Hague
(1964)
16. Schutz, A.: Collected Papers 5: Phenomenology and the Social Sciences. In: Embree, L. (ed.).
Springer, Dordrecht (2011)
17. Gilbert, M.: On Social Facts. Princeton University Press, Princeton (1992)
Collective Phronesis? An Investigation
of Collective Judgement and Professional
Action
Abstract This paper is a conceptual investigation of whether, and how, action con-
taining judgement can be understood as collective. Using an empirical example from
the police profession, the article tries to capture the complexity of a real situation
that requires both agency and judgement. The example is analysed in relation to
the idea of collective judgement, both regarding ongoing debates in phenomenol-
ogy concerning the possibility/need for a we-subject, and in social epistemology
concerning the idea of judgement based on social evidence. The article claims that
a richer account of action is needed in order to describe the collective aspects of
judgements. Drawing on an Aristotelian understanding of action, the paper argues
that an understanding of certain collective actions requires a notion of judgement
that involves both we-subjectivity, and social evidence.
1 Introduction
With examples drawn from the context of professional action, this paper will discuss
whether, and how, action containing judgement can be understood as collective.
Professional action is of especial interest with respect to the broader question of
collective agency since acting in one’s capacity as a professional often transcends
the individual horizon of an agent in different ways; for example, as professionals,
individuals represent a professional corps, meaning that they often act in relation
to situations requiring the use of specific skills that define a professionally trained
group of persons. The question is how this can be understood.
In book six of his Nicomachean Ethics Aristotle famously introduces several forms
of knowledge, one of which is “practical knowledge”. He goes on to distinguish two
Collective Phronesis? An Investigation of Collective … 25
different forms of human occupations or activities that require two different spheres
of practical knowledge, namely poiesis and praxis. While poiesis is production,
making or work, praxis is what we usually call action. Poiesis or work is, according
to Aristotle, characterized by producing something in light of a pre-assigned goal.
Work in this sense means producing a new entity that mimics a predefined product.
Whether it is a good entity, or not, is determined by the product’s definition, i.e. by the
goal formulated in advance of production. The skill or capacity required for work is
techne, art. According to Aristotle, art or techne, that is, the very process of making,
does not demand any overview of the context or the purpose of the product. “Art
is identical with a state of capacity to make, involving a true course of reasoning”
[4]. Techne is the skill of know-how. This very process is often seen as monotonous
or mechanical. It is not dependent on the fact a particular person is performing the
task, only that the agent possesses the requisite skills. It could, as Aristotle claims,
be performed by a slave. In the case of poiesis the very process sometimes requires
several agents working together on a shared goal or product, one example of this
might be a group of ancient boat constructers merging the logs in order to form the
hull of a boat. In terms of poiesis, collectivity does not seem so hard to grasp.
Praxis, or action, on the other hand, is introduced by Aristotle as the sphere of
interaction between humans. It is a mode of engagement in which the very process
is a goal where the goal is not external to the process. The implication here, though,
is that the individuals involved in praxis have a different relation to the very process.
It is not only the product that is valued, but also whether the ongoing actions are
performed in a good or skillful manner. In Politics Aristotle compares praxis with
life: “Life is action and not production” [4]. In life, and in praxis, humans do things
out of habit. Yet, when something unexpected or new appears, the imperative is
that habits and sedimented behaviours are changed. Praxis or action, according to
Aristotle, requires phronesis or practical wisdom. While techne is apparent within the
product (like a quality of the bread or the manufactured product) this is not the case
with phronesis. Phronesis shows itself in the very action, e.g. a surgeon’s capacity
to tell a husband that his wife did not survive the operation or a teacher’s ability to
convince a reluctant pupil to engage in her maths classes.
According to Aristotle, phronesis is always related to a certain action carried
out by a specific person. It shows itself in the very doing. In modern language, we
often try to approach this ability in terms of judgement. This gives us a good picture
of the kind of knowledge to which Aristotle aspires. But, at the same time, we
must be aware that, when talking about phronesis, Aristotle has in mind something
that is much wider than epistemic judgement. It is not only the very capacity to
judge in a particular situation, it also involves handling and coping with the situation
as a whole. Take the example of a surgeon; she shows her phronesis not only in
reaching a certain judgement during a complicated operation, but, rather it is present
throughout the process as a whole, in her way of using her art in order to prepare
for certain moments, to remain attentive to the possibility of potential complications
and so forth. Phronesis tends to reveal itself in unforeseen situations. One example
of this might be if during an operation, a surgeon suddenly runs into a complication
such that the patient’s life is in danger and is thus under pressure to make a crucial
26 E. Schwarz and J. H. Lappalainen
decision how to proceed. At these critical points, the surgeon’s confidence in her
own knowledge and (technical) skills, information about the patient, as well as an
overview of the medical resources available, forms the basis for the judgement she
must make; the quality of the judgement made in this critical situation will show
whether she possesses phronetic knowledge. Phronesis is “knowing how”, but in a
different sense to techne. The former is related to who the performer is—that is, with
respect to her educational history, previous experiences, her understanding as well
as the present interpersonal situation, e.g. the needs of the specific patient. Phronesis
is always related to others. Moreover, while techne or art is displayed in products,
phronetic knowledge is displayed in the actions of a particular individual, including
her judgements. Since phronesis is understood as something carried by a certain
person, it seems harder to understand it in collective terms, even if it takes place in
the interpersonal sphere of praxis.
Our claim is that phronesis might be understood as something that concerns the
collective. Thought in this way phronetic knowledge can be said to be displayed
in actions that encompass the judgements of a group of agents irreducible to the
judgements of individual group-members. For example, if a group of physicians co-
performing a complicated surgical operation, it is the group of agents that has to
act phronetically. It is not enough for one individual to act in a way that displays
practical wisdom based on previous experiences. We can even ascribe to the group
a shared history, along with possible joint experiences and established interpersonal
relations.
In a basic sense, we understand an activity as collective when it is dependent
on several agents. In some situations, as we have argued above, more than one
person is required to perform an action. This is the case, for example, when a group
of police officers takes control over a riot. In other situations, it may be that an
ongoing action is handed over to, or supplemented by, another. Examples of this
are shift changes in healthcare or a team of teachers that seek to teach a group
of pupils about the parliamentary system in an inter-disciplinary way. One way of
explaining the collective dimension of these activities would be to understand the
individuals involved as trained in a shared strategy and invested in a common goal.
For example, the police’s ability to handle a riot could be the result of joint training
as well as following distinct rules of engagement. The healthcare staff or teacher
team might be governed by common routines and jointly agreed goals and learning-
outputs. Here the very possibility of forming a collective is based on the fact that
persons share a certain formal knowledge, agreed rules or technical skills (techne).
But these activities are also part of an interpersonal praxis that requires not only
skillful interaction towards a predefined goal but also judgement regarding the very
process, experience and context of action [5]. Our claim is thus that these are activities
in the sphere of praxis, for example when a group of policemen (implicitly) judge
whether they are facing the possibility of a riot or just sporadic aggression shown by
particular individuals. It is therefore our view that this activity is to be understood as
operating within the sphere of praxis and therefore as activities that call for a form
of collective phronesis [12].
Collective Phronesis? An Investigation of Collective … 27
Within social epistemology there exists an ongoing debate about the possibility of
collective agency. In this debate, one often speaks of (at least) two distinct under-
standings of collective action, both of which are tied to the question of epistemic
judgement. In the first understanding, an individual must act upon a situation that
requires judgement based on what is called “social evidence”, i.e. the judgements of
others. Here, what is regarded as collective is not necessarily the concrete purported
act of judgement, but the character of the evidence. As stated above, in the field of
professional action, this often applies when different professions must collaborate
to judge upon something, e.g. in the event of considering and deciding on a clinical
diagnosis, when one physician will make a judgement on the basis of a range of
expertise from other experts.
The second understanding of collective judgement considers situations where a
group of individuals act in concert with respect to a joint goal, and doing so in such
a way that seems to imply a common judgement. This act of judgement itself pre-
supposes a form of agency that is irreducible to individual agents. Here the debate
broaches the question whether an ontological assumption of “group subjects” is
implied. On this, a variety of positions are adopted in the literature: eliminativism
is an ontological position rejecting the idea of collective judgements by a “group
agent”; to talk about a collective agent is to do so only metaphorically. Instead,
“anything ascribed to a group (…) can be re-expressed by reference to its members”
(List/Pettit 2011, 3). A similar position, singularism, which is common in economic
theory, criticizes group agency from a methodological perspective: according to this
position, the introduction of “group agents” would imply the unnecessary introduc-
tion of an entity to social ontology that is methodologically inaccessible. We cannot
1 We will not further discuss the role of discourse as a presupposition for an understanding of We-
Intentionality as it is e.g. discussed in relation to a Habermasian perspective, and as s expressed
through the use of language in different social contexts (see [6]).
28 E. Schwarz and J. H. Lappalainen
investigate collective acts without first relying on an analysis of individuals and their
judgements. A further view, “non-eliminative individualism”, held by List and Pet-
tit, combines both the abovementioned positions: Some uses of group agencies are
indeed metaphorical, “but often the notion of a “group agent” expresses a correct
observation” (ibid. s. 4). For List and Pettit, we need to analyse judgements made by
a group as if they had been made by individuals (methodological individualism) in
order do justice to the social world.
We cannot go any deeper into this quite complex and extensive debate. What
is interesting for us is not the different positions adopted and solutions suggested
to the problem of group agency and social evidence but the very way the problem
of collectivity is understood in the first place. In this sense, we share List/Pettit’s
methodological approach. Both understandings of collective action—based either
on social evidence or group subjects—are stated from a position exterior to the
group that aims at a certain goal or object. Even “methodological individualism”
treats action as a goal directed activity based on a judgement that can be described
on the basis of an observation.2 Yet we can imagine situations where a group of
agents, from an outside perspective are jointly related to a common goal and yet still
do not act as a group. They act without an understanding of what it means to judge
together, they lack a notion of each other’s position in relation to the situation in
which they are involved. If we would ask a member of the “group”, she or he would
not be conscious about being part of it. We can equally imagine ourselves being part
of a collective that might do things together, sharing certain activities, but all the
same we would not necessarily relate to such situations in a collective way. Instead
we just follow the rules or habits of the group, doing the things the group usually
does or is supposed to do [6]. By taking our point of departure in examples from
within the everyday practice of professionals one can see that questions about the
collective are complex in a way that seems not to be fully captured by the different
theoretical approaches we introduced above. In professional practice these different
cases of collectivity tend to intermingle. Let us take a closer look at an example from
a professional practice.
4 A Night Shift
Our example is taken from an essay written by a police student, who gives a personal
account of an experience from his work placement. As part of this account, the
police trainee analyses his own reflections from different theoretical perspectives.3
2 List and Pettit place emphasis on their commitment to realism, that is, to the world as an observable
and describable object independent from the constitutive perspective of any subject (ontological
realism) (ibid. 26). At the same time, the “methodological individualist” is interested in contrasting
“the methodological significance in detecting when agency is present, and when not” (ibid. 11).
3 This way of writing the final undergraduate paper is a method developed at the Centre for Studies
in Practical Knowledge. This specific essay-form is used at the Police education at Södertörn
University in Stockholm, Sweden, as well as at the Police University College in Bodø, Norway.
Collective Phronesis? An Investigation of Collective … 29
In our own analysis of the text, we will focus exclusively on the first part of his
essay. As a part of Swedish police education, students spend six months at a work
placement; during this period students are asked to work as police officers together
with their instructors and supervisors. The instructors are experienced police officers
whose responsibility is to introduce the student to practical police work, while the
supervisors are responsible for the students’ professional development. The student
describes his experience from a first-person perspective. Since we are dealing with a
beginner’s perspective here, we are afforded especially interesting insights into how
a trainee reflects on and questions established practices that more experienced police
officers might otherwise take for granted. The essay also displays the police student’s
desire to be accepted as a fully-fledged police officer. For these reasons, this inside
perspective makes the material interesting in light of the issue of the collective and
collectivity. The student wants to be part of the collective yet at the same time is
uncertain of what this collective consists. The example we are drawing upon also is
nonetheless limited, since we have access only to one person’s impression, and thus
nether to the others in the group nor to the group.
The example describes an emotionally demanding police action that potentially
could have turned into a violent and dangerous situation. This accentuates the ques-
tion of phronesis. Situations involving a latent threat of violence, and thus possibly
can change its meaning, require professional judgement. An important part of the
police’s professional knowledge is about acquiring an ability to read a situation,
judging quickly with which situation they are dealing, and thus knowing how to
act appropriately. It requires an ability to detect and manage threats of violence but
also to curb or respond to violence. Often these are situations where more than one
person is required, which is why there seems to be a certain requirement not only for
coherence in the reading of the situation but also a demand for mutual trust between
colleagues, both when it comes to reading possible threats of violence as well as
the ability to intervene and respond to any form of violence that may arise. This has
been described in research on police practices.4 The practical knowledge required
from police officers in situations defined by risks of strong affections and violence
is a capacity to sense emotions, moods and signs of affection in different individuals
[10, 11], but it is also a matter of communicating and sensing when, how and to
what extent authority or violence should be used [1, 10, 13]. These situations often
give limited or no space for interpersonal deliberation or discussion. As Knutsson
and Granér claim, it is impossible to have opinions on a colleague’s actions in an
ongoing violent situation [9]. The situation in the example below never turns into
a state of violence. Nonetheless, it does contain a risk of emotional volatility and
violence to which the police officers must relate.
The night shift starts with a meeting. The group leader describes the area where the oper-
ation will take place and what spots we will focus on during the night. Some places are
more interesting for us than others, spots where drug-dealing and drug-abuse are common.
4 Another crucial aspect it the professional capability to not provoke disputing persons to use violence
but instead to prevent the situation from becoming violent. This is discussed by the police officer
Markus Antonsson in an interesting text about how to communicate with groups of adolescents [3].
30 E. Schwarz and J. H. Lappalainen
The group leader gives instructions of the places we shall focus on, such as car parks and
basements. He tells us that we are supposed to work very “aggressively”. We shall proceed
“as usual”, seeking out, and frisk searching, suspicious-looking individuals.
Everyone sitting around the table seems to understand exactly what he is talking about,
nodding affirmatively. To me everything sounds a bit vague, and I start thinking about what
I consider to be “suspicious,” and if this might be the same as my colleagues take it to be. I
wonder […] if I can live up to my colleagues’ expectations, what they expect me to know,
and I especially wonder if I will manage to live up to my instructor’s expectations.
In the above excerpt we find some interesting aspects about the police trainee’s
views on collectivity and the collective. He claims that the instructions he receives
seem to presuppose knowledge he does not yet possess, and he expresses uncertainty
about, for example, how he might determine whether or not an individual is sus-
picious. We also see that he is anxious about not living up to his colleagues’ and
instructor’s expectations. He thereby shows that he has an idea, or a sense (at least)
of some kind of existing collective. His insecurity could be described as a feeling of
inadequacy in relation to the knowledge and attitudes he ascribes to this idea of the
collective. What is obvious is that he aspires to be a full member of the collective.
Further in his essay the student describes how his squad is instructed to investigate
an indoor car park often used by young people and adults to smoke cannabis. He
describes how his group organizes the search and closes off possible escape routes.
Thereafter he gives a vivid picture of the actual search in the car park that turns out
to be dark and thus difficult to survey in its entirety. After investigating this uneasy
environment for a while, they run into two men:
In front of the back door, not more than one meter away from where my colleague is waiting
at the other side, two people are sitting on red beer crates. One of them is wearing a suit,
which immediately makes me wonder what he is doing here. The other person is wearing
sportswear. I think to myself, biasedly, that [the second man] seems to be the sort of person
one expects to meet down here. Neither of the two men make a big fuss when we show up.
I show them my police badge and start talking to the two men while my instructor lets our
colleague in from the backside. In a nearby trash bin I can see a packet of rizzla papers. This
is a very common product used by cannabis-smokers. I am reflecting on the rizzla papers,
the location, and the time, and whether these factors suffice to establish reasonable suspicion
that would warrant a frisk search.
In this part of the squad’s activity we can see that the ongoing police operation is
essentially collective. The building could not have been both searched and secured
by one individual solely. In a potentially dangerous situation they have to rely on
the fact that other persons are securing their part of the area and also that they
provide adequate assistance in case of an outbreak of violence. Also in the situation
described above, the officers approach the two men and in doing so are expected to
perform different tasks as the situation requires; the police officers are thus acting
interdependently. They are acting so to say as “one body”. This means that neither
one single policeman is able to handle the situation on his own, nor is the “operation”
reducible to a cluster of individual actions performed by individual agents.
We also see that the police student is attentive to whatever he discovers or sees.
Further he reflects on his impressions and what might be expected of him. He displays
Collective Phronesis? An Investigation of Collective … 31
a strong attentiveness to the situation, its potential dangers and at the same time to
his colleagues’ actions and reactions. He relates to and reflects on numerous aspects:
the group, the situation, his ideas of his colleagues’ expectations. This becomes even
more evident in the following passage:
Both men are very cooperative and do precisely what we say. I think that it is a bit strange
that neither of them shows any signs of drug use. Can it be that I am missing obvious signs,
or are they just not affected by drugs at all? While my colleague and I look after the men,
my instructor is checking their ID over the phone. He tells us to come over to see what he
has found: absolutely nothing. Neither man shows up in the police register. My thoughts
are: “Ok, that’s good. These are nice guys.” My instructor asks me what I think we ought
to do with them. My mind is spinning: They show no signs of drug use, which speaks to
their innocence. What works against them is that there is rizzla papers close by. The time,
as well as the place is also suspicious. I wonder what my instructor will think about me
and my professional knowledge if I give the wrong answer. This makes it difficult for me to
decide. I remember how earlier this evening the group leader told us that we should work
“aggressively.” So, even if I don’t think we will find any drugs, I say: “Let’s frisk search
them.”
Again, in this quote the collective aspects of the activity become evident. Two
police officers “look after the men” while one officer checks their criminal records.
The division of tasks does not seem to be preceded by any form of dialogue. It
seems as if they already have a common understanding about how the division is
to be made. This is an example of how they collectively maintain their power and
superiority in this public location that now has turned into their working place. Once
more we get the impression of them acting as one body. Then comes the moment when
the investigation of the location and the men reaches a critical point; the squad has
searched the area, confronted the two individuals, made an impression of them, talked
to them and received information about their potential criminal records. Precisely
at this point the squad must decide how to proceed. This is the moment where we
can see that the situation seems to involve a judgement (based on the preceding
moments in the whole operation), namely whether the two men ought to be labelled
as suspicious or not.
We see this moment actualised by the instructor’s question to the student. He asks
the student if he thinks that the men ought to be frisk searched. The very fact that the
instructor poses this question shows that a judgement is required. We see from the
student’s description that he perceives this as a test of his personal ability to judge.
The police student then describes his own ambivalence: on the one hand, he does not
believe there are sufficient grounds for a frisk search but, on the other, he runs into
doubts about his capability to read the situation correctly. We also see how important
it is for the student to appear as both a competent and a full member of the squad. In
his description, he seems to have come to the conclusion that the other officers think
that the men ought to be frisk searched, and finally he concludes that the men should
be searched. Unsurprising, since he writes that he hopes to live up to his colleagues’
expectations. This is nonetheless interesting; his account implies that while he has
an impression that the others concluded that the men should be frisk searched, he
did not. He thus ends up expressing what he believes to be the conclusions of the
collective.
32 E. Schwarz and J. H. Lappalainen
The two men are thereafter searched and this very act seems to imply the judge-
ment that the two men were acting suspiciously. We cannot know for sure that the
act of frisk searching the two men is based on a judgement in an Aristotelian sense.
However, what we can see is that if it were, it would be a collective judgement differ-
ing from (at least) one of the individual judgements in the situation. If a judgement
takes place, it is not reducible to the sum of the individuals’ judgements.
5 Collective Phronesis?
the two men. If the operation leading up to and including the frisk search is to be
regarded as an action in the Aristotelian sense, it implies that the men were judged
as suspicious individuals. It is through the judgement that the action (praxis) is
distinguished from goal-directed work (poiesis). In an Aristotelian understanding
of action, the judgement cannot be reduced to a conclusion based on intentions and
goals set up in advance. In praxis, phronesis might emerge when the individual comes
to a judgement based on her knowledge and understanding of the actual situation.
Judgement is in this sense essentially a trial. In order for a judgement to be collective
this trial ought to imply an interpersonal aspect. What we know from this situation
is solely the police student’s description of the dialogue with his supervisor and his
own sprawling thoughts on the issue. We have no way of knowing how the other two
policemen thought or to what extent they were considering the situation. We thus
cannot determine how they came to their own conclusions that serve to comprise the
collective judgement. What we do know is that the police student’s answer to his
instructor is not an expression of his own judgement of the situation, but rather an
effort to fulfill a predefined task. The trial to which he responds is something other
than the trial of the collective. If we get back to the distinction made by Aristotle
between poiesis and practice, work and action, then, as a novice, the student displays
that he is not in the sphere of praxis but in the sphere of poiesis. In an Aristotelian
sense, he is thus not acting, he is just doing his work trying to accomplish his tasks,
to fulfil a predefined goal. His answer to his instructor shows that his main purpose
is to do his work as a good student. It is the product (i.e. giving either the right or
wrong answer) that is the focus of his decision.
We have to acknowledge that it is not possible to know for sure if this collective
activity can properly be described as action in Aristotelian terms at all. It might be
the case that the two other police officers were just trying to fulfil the group leader’s
order and predefined tasks, and thus were not acting in a phronetic way at all. It might
also be possible that the two other police officers ended up with identical individual
judgements. Therefore, it is inconclusive whether their activity is an example of
poiesis or praxis. However, by highlighting the student’s trial, the actual meeting
between the individuals’ perspectives becomes evident. Here at least there is space
for the possibility of the formation of collective judgement.
For us, operating within an Aristotelean framework, there are at least three aspects
that are important about the very possibility of collective phronesis: First, in an Aris-
totelian understanding of action, the judgement is something attempted and formed
in the act itself. Action entails intervening in an uncertain future, and this always
involves an element of risk, which manifests in the very act of passing judgement.
And if we would like to understand the situation as an action containing phronesis
and thus a judgement, it is apparent that the collective arrives at a judgement not in
the same way as individual persons. Here we run into a situation in which an action
containing a collective judgement cannot be explained with reference to participating
individuals and their single judgements.
Secondly, if the example we have described above is understood in terms of an
Aristotelian conception of action then the police officer’s collective activity is shown
in its complexity. Of interest, here is that the example from police practice shows that
34 E. Schwarz and J. H. Lappalainen
collective judgement is not a cluster of individual judgements but rather the result
of a dynamic process. Collective action and judgement both emerge from within a
plurality. This plurality might be explained as the police student not being a part of the
collective. But it might also be explained by the fact that a collective is dynamic in the
sense that the individuals involved possess different levels of experience, knowledge
and goals. These different levels might be a part of the judgement.
A third aspect broaches methodology, specifically a question about the relation
between different perspectives that arise within the situation. We see that none of
the presented ideas of collective action suffice to explain the collective act in this
example. The collective action cannot be explained solely as an instance of social
evidence, i.e., as a judgement based on the sum of plural judgements. The judgement
of the police student might be said to rely on social evidence (or rather on his wish for
their experience as deeper knowledge), but this is not the whole story about how he
makes his judgement. His judgement is rather determined by an aspiration to judge
in line with his colleagues, and thus his decision seems to be strongly influenced by
his wish to be a part of the collective. This becomes clear once we have accounted
for the inside perspective of the student. From the outside perspective, we see a
collective action that tends to imply a judgement. In that sense the action seems to be
an example of group-agency. The fact that they decide to frisk search suggests that
the squad has come to a collective judgement surrounding the suspicious men. We
know that the student did not arrive at this judgement on his own. But as a collective
this is the judgement implied in the action. The situation thus contains both group-
agency and an idea of the importance of social evidence, as well as an emotional
understanding of the collectivity.
One difficulty with our example is that it can be interpreted in several ways. It
could be interpreted as an example of corrupt police work in a gratuitous exercise of
power. But it could also be interpreted as an example of the police student’s inability
to detect or notice facets of the situation that were obvious for the more experienced
police officers. The empirical material we have analysed from the police student is
insufficient to determine whether the action is phronetic or ought to be understood
as an example of Aristotelian praxis. Nonetheless, what we have here is an example
of an act containing a judgement that is irreducibly collective, or it is not as an action
at all, i.e. it is an example of work with a predefined goal.
6 Concluding Reflection
We would thus claim that by taking our point of departure in an actual example from a
professional practice, action understood in terms of Aristotelian praxis might appear
as irreducibly collective. We further claim that neither the framework offered by
social epistemology nor by phenomenology can sufficiently explain how action that
includes judgement can be considered collective. We suggest that an Aristotelian
understanding of collective action must consider the dynamics of collectivity. A
collective consists of members with different skills, different levels of knowledge,
Collective Phronesis? An Investigation of Collective … 35
References
1. Algård, J.: Det är inte olagligt att ta livet av sig [It is not illegal to commit suicide]. In: Ekstrom
von Essen, U. (ed.) Den goda polisen. Sju essäer om reflekterad yrkeserfarenhet. [The Virtuous
Police. Seven essays on reflected professional experience]. PHS serie i polisiärt arbete, Solna
(2008)
2. Annas, J.: Intelligent Virtue. Oxford University Press (2011)
3. Antonsson, M.: En höna av en fjäder [Mountain of a molehill]. In: Ekstrom von Essen, U.
(ed.) Den goda polisen. Sju essäer om reflekterad yrkeserfarenhet. [The Virtuous Police. Seven
essays on reflected professional experience]. PHS serie i polisiärt arbete, Solna (2008)
4. Aristotle: The Nicomachean Ethics. Harvard University Press, Cambridge, Mass (1975)
5. Collins, H.: Tacit and Explicit Knowledge. University of Chicago press, Chicago/London
(2015)
6. Giovagnioli, R.: From habits to rituals: rituals as social habits. Open Inf. Sci. 181–188 (2018)
7. Goldman, A.: Knowledge in a Social World. Oxford University Press, Oxford (1999)
8. Goldman, A.: Experts. which ones should you trust ?. Philos. Phenomenological Res. 63(1),
85–110 (2011)
9. Granér, R., Knutsson, M.: Perspektiv på polisetik [Perspectives on Police Ethics]. Studentlit-
teratur, Lund (2001)
10. Heder, M.: Vad är en bra polis? [What is a Good Police Officer?]. In: Vad är praktisk kunskap?
[What is Practical Knowledge?], Södertörn Studies in Practical Knowledge, Huddinge (2011)
11. Hoel, L.: Politiarbeid i praxis [Police Work in Practice]. Universitetsforlaget, Oslo (2013)
12. Kinsella, E., Kinsella, A.: Phronesis as Professional Knowledge. Sense Publishers, Rotterdam
(2012)
13. Krogh, U.: “Krona eller Klave”. In: Ekstrom von Essen, U. (ed.) Den goda polisen. Sju essäer
om reflekterad yrkeserfarenhet. [The Virtuous Police. Seven essays on reflected professional
experience]. PHS serie i polisiärt arbete, Solna (2008)
14. Krueger, J.: Phenom. Cogn. Sci. 13, 533. (2014). https://doi.org/10.1007/s11097-014-9363-1
15. List, C., Pettit, P.: Group Agency. The possibility, design, and status of corporate agents. Oxford
University Press (2013)
16. Szanto, T., Moran, D.: Phenomenology of Sociality. New York, Routledge (2017)
17. Von Brandenstein, BH.: Handbuch der Philosophischen Grundbegriffe, Bd. 2, München (1973)
From Searle’s Speech Acts to Smith’s
Truth-Makers
Philip Larrey
Abstract Beginning with Searle’s meaning theory and his account of mental acts
(“intentionality”), the article identifies the limits of Searle’s approach from the per-
spective of common-sense reasoning. The second part explores the notion of “truth
makers” in the work of Barry Smith, inasmuch as it can be seen to complete a robust
meaning theory in the analytical tradition.
1 Introduction
P. Larrey (B)
Pontifical Lateran University, Vatican City 00120, Italy
e-mail: larrey@pul.va
mode of speech. So if, prior to using the name, we came to an agreement on the
precise characteristics which constituted the identity of Aristotle, our rules for using
the name would be precise”.1
For most of the twentieth century in analytic philosophy, meanings were banned
or at least looked down upon. Quine’s lasting criticism of essentialism (where mean-
ings separated from words become like Platonic essences separated from individual
beings) ushered in many decades of avoiding meanings and longing for a purely
extensional logic (because meanings were only operable within intensional log-
ics). With Donald Davidson, we once again are able to speak about meanings and
intensional content. Hilary Putnam’s quite famous article, Meaning of Meaning, [2]
brought us full round as he maintains that the prephilosophical notion of mean-
ing is almost completely mistaken. But even Putnam’s criticism of the “everyday’s
understanding of meaning” has not held up entirely.
2 Meaning
A prephilosophical notion of meaning still has its merits. When we look at how
effective body language can be at transmitting meanings (independently of language
per sé) [3], we can conclude that there must be something that is at work between two
interlocutors who do not speak. Social Engineer, Inc., specializes in helping com-
panies realize how information can be “stolen” or received by hackers through non-
verbal communication (“… malicious hackers are able to exploit human weaknesses
to obtain access to information and resources through manipulation and deceit”2 ).
Recent hacks on Jeff Bezos’ cell phone,3 Sony,4 HB Gary5 and Lockheed Mar-
tin6 all have non-digital components involved, which indicates that to some extent,
meanings (and even information) can be separated from words and/or sentences. We
can also recall how the Stuxnet virus was uploaded into the computer system at a
nuclear refining facility in Iran by the U.S. and Israeli military. This malware was a
destructive hack which ruined many of the centrifuges that the Iranians were using
to produce enriched uranium. It infected the system because a plant worker found a
1 Ibid.,172.
2 See Ibid., xvi.
3 See https://www.businessinsider.com/theories-on-how-national-enquirer-got-jeff-bezos-private-
photos-2019-2.
4 See https://www.washingtonpost.com/news/the-switch/wp/2014/12/18/the-sony-pictures-hack-
explained/?utm_term=.c0caf75333d3.
5 See https://arstechnica.com/tech-policy/2012/03/the-hbgary-saga-nears-its-end/. The saga nears
its end because almost all of the hackers were caught and put in jail.
6 See https://www.computerweekly.com/opinion/No-excuses-for-Lockheed-Martin-cyber-attack.
Even more recently, it is quite apparent that the Chinese continue to get access to sensitive content
from the Defense Contractor. When we examine the Chinese new J-20, the resemblance to the U.
S. F-22 is striking. See https://www.cnbc.com/2017/11/08/chinese-theft-of-sensitive-us-military-
technology-still-huge-problem.html.
From Searle’s Speech Acts to Smith’s Truth-Makers 39
pen drive in the parking lot and unsuspectingly plugged it into his computer station.
The very digital hack began in a very non digital fashion.
Slightly less philosophical could be the uncanny way that dogs (as well as other
domestic pets) can actually grasp (in some non-linguistic sense) the meaning of their
owner, without understanding language. How does the dog realize what the owner
wants or how the owner feels, even though it has no access to linguistic structure?
How can dogs react also to similar body-language signs in order to fetch, come, stay,
or lay down? The emotional bond between certain types of dogs and their owners
testifies to the presence of intensional content which is independent of normal verbal
language.
On Searle’s account of meaning,7 we have brain states and mental states which
coincide. The brain causes meaning by causing mental states, yet there is really
nothing more to the mental state than the physical and chemical states in the brain
itself. Because of this approach, Searle is often classified as a reductionist, reducing
the mental (or spiritual) to the physical components of the brain (cortex).
How do mental states get transmitted to our interlocutor? Usually through lan-
guage. That is why each must understand the language of the other in order to convey
“mental entities” to one another. If you understand the language of the other, this
happens naturally and easily. If not, it is more difficult but still possible.
Is it possible to communicate mental states without language? Yes, as we have
mentioned, through body language, pointing, sounds, and so forth. If those ges-
tures are considered language, then the point is moot. However, the prephilosophical
notion of meaning would seem to be defended, as the transmitting of meaningful
mental states between interlocutors (even in the case of the dog who comes and goes
according to the wishes of the owner).
From a neurobiological point of view, all we can go on is the brain state, which we
can map quite successfully and even interpret what is happening in the brain. We can
take for example the work of Robert Knight and his colleagues at the University of
California, Berkeley (Neuroscience Lab). He writes, “We recorded electrical signals
directly from the human language areas when a person heard words. We then decoded
these electrical signals and were able to turn them into sound files that reflected what
the person heard, with remarkable accuracy”.8 Remarkably, the team was then able to
decode speech when a person thinks of a specific word, from direct brain recordings.
“The new techniques and mathematical processing of the brain signals got us closer
to the details we need to extract the signals that are relevant for reproducing speech”.9
They are close to creating a prosthetic implant which can reproduce words on the
basis of the neurological activity in a person who is unable to speak.
7 Above all, that found in his book, Intentionality. As Essay in the Philosophy of Mind. Also, his
short work, The Mystery of Consciousness, his The Rediscovery of the Mind, and simply, Mind. A
Brief Introduction.
8 See https://blog.frontiersin.org/2016/05/27/setting-free-the-words-trapped-in-our-heads/, by
Monica Favre, Frontiers Science writer.
9 Ibid.
40 P. Larrey
With just a little imagination, we can extrapolate the example and discover the
following: a brain-to-brain connection which bypasses speech all together, and trans-
mits the appropriate neurological activity between two interlocutors without the need
to use spoken language.
Although this may sound like science fiction, it has already been tried. At the
University of Washington, Seattle, neurologists were able to establish direct brain-
to-brain links, asking each participant a series of questions (without spoken language)
and testing the results.10 The accuracy of correct answers was 90%. In addition, Elon
Musk wants to have us all connected directly and has founded a company for that
purpose, Neuralink, and has raised $27 million to fund it.11
A crucial distinction needs to be made at this point. Neurological activities and
their patterns are not the same as thoughts. Thoughts are logical entities that are
expressed through words, and the brain states that correlate to thoughts are physical
processes and can therefore be mapped and transmitted through an implant. There
is nothing science-fictional about that. But achieving that (which is a wonderful
thing for people who cannot communicate—think, for example, of the late Stephen
Hawking) is not the same as achieving telepathy. If we do not reduce mental states to
brain states, then we conclude that the brain state is something more and immaterial.
3 Truth Makers
10 See http://www.washington.edu/news/2015/09/23/uw-team-links-two-human-brains-for-
question-and-answer-experiment/.
11 See http://www.businessinsider.com/elon-musk-neuralink-raises-27-million-2017-8. This is
Musk’s fourth company, and although the idea sounds absurd, people tend to take Musk seriously.
From Searle’s Speech Acts to Smith’s Truth-Makers 41
utterances whose truth value can change over time. Some critics of Quine have even
suggested that substituting eternal sentences for propositions turns out to be almost
the same thing.
From the point of view of the history of analytical philosophy, eliminating propo-
sitions as truth vehicles required obtaining others. Barry Smith then suggested the
concept of truth-makers. “The fall from favour of logical realism brought with it a
corresponding decline of interest in the ontology of truth. The notions of correspon-
dence and indeed of truth itself first of all came to appear obscure and ‘metaphysical’.
Then Tarski’s work, while rehabilitating the idea of truth, seemed to embody a rejec-
tion of a full-blooded correspondence. (3) In the wake of Tarski, philosophers and
logicians have largely turned their attentions away from the complex and bewildering
difficulties of the relations between language and the real world, turning instead to
the investigation of more tractable set-theoretic surrogates”.12
The reason why truth-makers became important in the analytical tradition lies with
the inheritance left by Tarski. The influence of his article on the semantic concept of
truth cannot be overstated. Simplifying that influence somewhat, it can be argued that
two impressions were conveyed and affected almost all of subsequent trends in the
analytical tradition. The first impression was that in terms of a “correspondence the-
ory of truth”, Aristotle’s analysis suffices to understand the prephilosophical notion
of truth as the conformity between what is stated and what there is. This stance
becomes obvious from a common sense perspective: to say what is that it is, and
what is not that it is not, is to speak the truth. If it is raining outside where you are,
and you utter, “It is raining outside” means that the utterance is true. If it is not
raining, and you say, “It is not raining outside” means that the utterance is true as
well. With almost all examples we can examine, Aristotle’s description adequately
deals with what we intuitively understand as truth.
The second impression of Tarski’s game changing article is that the introduction
of a “coherence theory of truth” no longer deals with real-world truth relations, but
rather with truth for a formal language. In fact, the analytical tradition favored the
coherence theory of truth precisely because of the birth and development of formal
systems in the philosophy of logic. Although the formalization of language and logics
systems has its merit, it does not deal with real-world truth relations, and therefore
does not concern itself directly with that which matters to those who seek the truth
from a realist’s point of view.
Hilary Putnam concluded that “[I]f we want to account for truth, Tarski’s work
needs supplementing with a philosophically non-neutral correspondence theory”.13
12 Reference [5]. The footnote used in the original article refers to Aristotle: “3. Aristotle’s famous
To say of what is that it is not, or of what is not that it is, is false, while to say of what is that
it is, or of what is not that it is not, is true
(Met., 1011b 32 ff.) is, as Tarski himself is anxious to claim (1944, p. 343), less than full-blooded
correspondence theory, but Aristotle is elsewhere (op. cit., 1027b 22, 1051b 32 ff.) prepared to
speak of truth reflecting ‘combinations’ of subject and attribute in reality.”
13 Ibid., 288.
42 P. Larrey
Such is the task that Barry Smith and his colleagues set out to accomplish with
the notion of truth-makers. Truth makers are more than truth-bearers (or in Quine’s
vocabulary, truth vehicles) because they are those rational entities in virtue of which
sentences or statements are true. The truth vehicles simply carry the burden of the
truth predicate (i.e., “is true”), whereas the truth-makers are those logical constructs
that determine the truth value of propositions.
Barry Smith’s initial point has to do with what he calls “Moments”. Moments
are objects that require other objects in order to exist at all; they are not ontologi-
cally autonomous, but rather dependent on other objects for their existence. Some
interesting examples include a human smile (which requires a human face in order
to be) or a duel, which requires duelers in order to exist (although the “duel” is not
reducible to the duelers).
Such a distinction can also be traced back to Aristotle, with his now classical dis-
tinction between substantial (not in a subject) and accidental (in a subject). Although
there are nuances which differ between “moments” and “substances”, the notion is
close enough to be useful for the analysis of truth-makers. “Substantial” is not in a
subject, but rather the subject itself; while “accidents” adhere in a subject and are
therefore not substantial. Such a distinction runs through the entire scholastic period
(especially in Aquinas) as well as modern philosophy with Descartes, Locke and
Hume.
Smith summarizes: “[…] We shall here consider the theory obtained from the view
that what makes it true that Socrates died is Socrates’ death [example borrowed from
Russell], what makes it true that Amundsen flew to the pole is his flight [example
borrowed from Davidson], what makes it true that Mary is smiling is her (present)
smile, and so on. Or, in other words, that for many simple sentences about spatio-
temporal objects the truth-makers for these sentences are the moments picked out
by gerundials and other nominalized expressions closely related to the main verbs
of the sentences in question”.14
The notion of “gerundial” is crucial to the theory of what makes a sentence true.
Simply put, what makes something true is it’s being thus. In the above examples,
we can slightly modify the verbs by converting them into gerundials (or gerunds).
Thus, we have “what makes it true that Socrates died” is Socrates’ dying, not in
the sense of an ongoing action or happening (i.e., not in the sense of the process of
Socrates’ death—“dying”—but rather in the sense of “Socrates having died”). Smith
attempts to capture the action verb form in order to render the noun more dynamic
and therefore more clearly a “maker” than just a truth vehicle.
The notion of truth makers extends beyond the normal boundaries of both the
analytical tradition and neo-positivism in general. One of the central tenants of the
neo-positivist tradition was verificationism, where the meaning of a statement implied
understanding the conditions under which the statement was true. Even in the twen-
tieth century, such a theory came under fire and faced harsh criticism. One line of
criticism (still valid today) is that I can understand the meaning of a phrase without
knowing whether it be true or false. However, a valid response from the neo-positivist
14 Ibid., 296/7.
From Searle’s Speech Acts to Smith’s Truth-Makers 43
is that I cannot understand the meaning of an utterance without knowing what would
make it true or false, i.e., the truth conditions (from an empirical point of view).
Smith concludes: “A knowledge of truth-conditions takes us at most one step
towards reality: one can, surely, envisage understanding a sentence (knowing its
meaning), whilst at the same time having only partial knowledge of the nature of its
possible truth-makers”.15 “Partial knowledge” signifies “given our current stage of
comprehension”. Due to the constant progress of scientific knowledge, our knowl-
edge is always partial compared to future knowledge. When Tiberius sat in his villa
on Capri Island in the first century and pointed to the bay of Naples describing such
waters as “magnificent and profound”, he had no idea that he was referring to H2 O.
His knowledge of chemistry did not get that far. Yet, what he meant and to what
he referred are very clear. Having only partial knowledge of what makes a sentence
true does not lead us to skepticism, but rather takes us down the path of a full-bodied
epistemology in the realist tradition. Even though today we can substitute H2 O for
“waters of the bay of Naples” and the statement comes out true, such a substitution
salva veritate was not possible for the Emperor two thousand years ago.
Aside from the neo-positivistic limitation in terms of meaning and truth, within the
analytical tradition the notion of conceptual schemes exerted an enormous amount
of influence in the second half of the twentieth century. Thanks not only to Quine,
but also Thomas Kuhn and others, truth became relative to one’s conceptual scheme.
Quine referred to this as “ontological relativity”, in which what there is is deter-
mined by the conceptual scheme. For Kuhn, who popularized the notion with his
term “paradigm” (and associated term “paradigm shift” describing how we abandon
one paradigm in favor of another), what there is is determined by the paradigm. If
reality is dependent on the conceptual scheme, so too are truth values (and even truth
conditions). As Donald Davidson writes in his classic essay, On the very idea of a
conceptual scheme16 “Reality itself is relative to a scheme: what counts as real in
one system may not in another”.
If reality is relative to a scheme or paradigm, certainly truth is also. What is
true in one scheme may not be in another. Although this relativity of truth can be
adjudicated within a realist tradition (where some truths are dependent on culture,
history or even linguistic communities), that which is completely objective is the
meaning of the truth predicate. What is true for a Chinese speaker may not be for
a Spanish speaker, but what “is true” means is the same for everyone. Such is the
objectivity of the truth predicate across all natural languages.
From the perspective of truth makers, Quine’s ontological relativity cannot be
sustained. Davidson’s devastating critique of conceptual schemes concludes with
one of his most famous paragraphs and sets up the need for a conversation concern-
ing truth-makers. Davidson writes: “Nothing, however, no thing, makes sentences
and theories true: not experience, not surface irritations, not the world, can make
a sentence true. That experience takes a certain course, that our skin is warmed or
punctured, that the universe is finite, these facts, if we like to talk that way, make
15 Ibid., 299.
16 See [6], pp. 183–198. The quoted line is on page 183.
44 P. Larrey
sentences and theories true. But this point is put better without mention of facts. The
sentence ‘My skin is warm’ is true if and only if my skin is warm. Here there is no
reference to a fact, a world, an experience, or piece of evidence”.17
Because of this paragraph, Davidson is often misunderstood. When he writes that
“Nothing, no thing makes sentences true”, he remains a realist in his conception
of truth. That which he attempts to critique is the effort to place an intermediary
between a speaker’s language and reality she is describing (i.e., a conceptual scheme,
a paradigm, a framework and so on). There is ‘no thing’ which renders a phrase true
aside from the truth-makers (as in Smith’s account), which simply put means that
reality is as described by the speaker. Davidson crystalizes the point: “In giving up
the dualism of scheme and world, we do not give up the world, but re-establish
unmediated touch with the familiar objects whose antics make our sentences and
opinions true or false”.18
In Davidson’s account, we do not need an epistemological construct which “maps”
language onto reality, or fixes language to some non-linguistic reality. The reason why
there is no need for such a construct is that language does not need to be “hooked on
to reality” (as Hilary Putnam would suggest) precisely because language arises from
reality as described by a speaker (of any natural language whatsoever). Davidson also
preserves objectivity by examining speakers of a language as a community and not
as solipsistic individuals only. Speakers of a given language understand each other
because with their words, they refer to the same reality (assuming they speak the same
language). This account is often referred to the “triangular approach” of Davidson,
where the three points of the triangle are the two speakers and the denotated reality,
already “hooked together” by a common language.
4 Conclusion
In conclusion, the analysis presented here traced the notion of meaning and intention-
ality in John Searle as pertinent for the examination of truth and the truth predicate.
By way of contrast with neo-positivism and the analytical tradition as a whole, Barry
Smith offers his notion of “truth-makers” which are not merely truth vehicles or
conditions of truthfulness, but rather that which makes sentences true or false. In
other words, using the gerundials, the being of that which is (in an empirical sense).
Furthermore, such an account can be seen to dovetail well with Davidson’s account
of the “triangular” nature of human natural language, where there is no need of an
epistemological or logical construct which would “hook” language onto reality (or
on to a “fact” or on to an “event”). The notion of truth-maker adequately makes
explicit that which is implied by the Davidsonian account.
17 Ibid., 194.
18 Ibid., 198.
From Searle’s Speech Acts to Smith’s Truth-Makers 45
References
1. Searle, J.R.: Speech Acts: an Essay in the Philosophy of Language. Cambridge University Press,
Cambridge, MA (1969/1994) (17 reprintings)
2. Putnam, H.: Meaning of meaning. Minnesota Stud. Philos. Sci. 131–193 (1975)
3. Hadnagy, C.: Unmasking the Social Engineer: the Human Element of Security. Wiley & Sons,
Indianapolis (2014)
4. Quine, W.V.O.: Word and Object. MIT Press, p. 170 (1960)
5. Smith, B., (University of Manchester), Mulligan, K., (University of Hamburg), Simons, P. (Uni-
versity of Salzburg): Truth makers. Philos. Phenomenological Research, vol. 44, pp. 287–321.
(1984)
6. Davidson, D.: Inquiries into Truth and Interpretation. Oxford University Press, Oxford (1984)
Science, Ethics and Social Practices
The Biological Logic of Human Action:
On the (Considerable) Difference
Between “Rational” and “Adaptive”
Douglas A. Marshall
Abstract Modern democratic and capitalist (i.e. classically liberal) societies are
constructed and justified on a foundational assumption that human action conforms
to a rationalist logic, approximating, if not identical to, that of instrumental ratio-
nality. This chapter argues that because human behavior has evolved to be adaptive,
it adheres to a very different, biological, logic. After establishing the multi-layered
reasons that human beings could not, and did not, evolve to be rational, it out-
lines the thoroughly social behavioral capacities and dispositions that could and did
evolve, which together constitute the actual adaptive logic of human action. In a
nutshell, the resultant model replaces the rationalist model of individualist, single-
minded, amoral, materially maximizing, informed, and calculating human actors
with an empirically and theoretically substantiable model of groupish, distractible,
moral, Satisfycing, epistemologically conforming, and habitual actors. The chapter
concludes with a consideration of several significant implications of the improved
model for individuals, for social science, and for social policy/governance.
1 Introduction
This chapter is distilled from a longer, unpublished, manuscript, which interested readers can soon
find, along with a comprehensive bibliography, at the author’s Academia.edu page.
D. A. Marshall (B)
University of South Alabama, Mobile, AL, USA
e-mail: dougmarshall@southalabama.edu
At about the same time that Durkheim was dismissing rationality as a foundation
of society, Max Weber was preoccupied with its implications for society. He saw
the perfusion of instrumental (“zweck-“) rationality throughout modern society and
culture as “the master trend of history” [1]. For him, and many subsequent observers,
the rational actor is:
Individualist. The rational actor is exclusively interested in their individual out-
comes. Those of others matter only to the extent that they impact the actor’s own.
Colloquially, the rationalist assumption is that individuals are selfish.
Single-Minded. The rational actor is fixated on a single goal to the exclusion of
all competing or ancillary ends. Accordingly, rational actors are commonly described
as amoral, since questions of right and wrong (like every other competing desiderata)
are irrelevant to them.
Maximizing. The rational actor is insatiable, in that for them, more is always
better. Like “right” and “wrong”, the concept of “enough” is meaningless, as their
appetite for their particular maximand is boundless.
The Biological Logic of Human Action: On the (Considerable) … 51
Materialist. In theory, the rational actor can seek to maximize any conceivable
end, but most applications gravitate towards material, usually financial, ends, given
the analytical need for quantifiable, calculable, outcomes.
Informed. The rational actor is presumed to have access to, and to avail themselves
of, any and all information of relevance to the pursuit of their goal.
Calculating. The rational actor uses that information to calculate the impact that
of each of the options available to them would have on their goal in order to select
the optimal means of maximizing it.
The best-known shortcomings of the rationalist assumption concern the fact that
the sheer amount of information and calculation it requires far exceeds human cog-
nitive abilities. Such findings gave rise to “bounded” models which stipulated that
while actors might not be able to obtain all relevant information or to calculate all
possible consequences, they will at least avail themselves of whatever information,
and perform whatever calculations, they can. But even this “thinner” rationality was
undermined by further discoveries that people regularly fail to make use of even
readily accessible information, or to engage in any more calculation than they abso-
lutely have to. We are, in short, “cognitive misers” [2]. Most damaging of all has
been evidence that human behavior diverges from rationality in ways irreducible to
either limits or laziness—that it is “predictably irrational” [3]. Thus, the primary
shortcoming of the rationalist logic of human action is that human beings behave
according to a different logic altogether.
This chapter first lays out the nested “negative” arguments as to why evolution
was highly unlikely to ever produce the capacity or dispositions necessary to an
inherently zweck-rational species, and then retraces those arguments to construct the
“positive” case outlining the biological logic of human action.
The justification for assimilating adaptivity to rationality goes like this: Individual
fitness is primarily a function of survival, and survival, in turn, is primarily a function
of maximizing material resources. Since rationality is by definition the optimal means
of maximizing material resources, and evolution invariably finds and follows optimal
solutions, organisms inevitably behave rationally. This chain of reasoning includes
at least three dubious assertions:
1. That evolution inevitably arrives at optimum solutions.
2. That survival is primarily a function of material resource maximization.
3. That individual fitness is primarily a function of survival.
Let us address each in turn.
52 D. A. Marshall
problem is that evolution can’t even “see” (or select) any single gene’s phenotype,
only entire organisms, the cumulative products of tens of thousands of interacting
genes.
Between the primacy of variation, the fact that evolution is a negative and ret-
rospective process, and its blurry view of genes, it is extraordinarily unlikely that
the specific collection of traits that together constitute rationality could ever have
evolved, even were rationality an optimally adaptive pattern of behavior—which it
isn’t.
The belief that instrumentally rational behavior necessarily optimizes fitness reflects
a fundamental misunderstanding of the adaptive problem faced by organisms, in that
it reduces fitness to a problem of material maximization—most often of calories. But
even the proximate aim of survival, much less the ultimate goal of fitness, has little
to do with material maximization. At best, rationality is orthogonal to (uncorrelated
with) survival, but more commonly antithetical to it. To see why, let us consider the
defining features of rationality on a case-by-case basis.
Individualism. The supposed survival value of individualism derives from a con-
ception of evolution wherein the primary competition for survival is that among
individuals. But interindividual competition is but one of many kinds of evolution-
ary challenge. Darwin helpfully clarifies the field by distinguishing between “abi-
otic” and “biotic” competition [5]. Abiotic competition refers to non-living things
threatening one’s fitness—gravity, storms, heatstroke, hypothermia, dehydration, and
starvation. Biotic competition (living threats to one’s fitness), further breaks down
into: Inter-specific competition—threats posed by members of other species, and;
Intra-specific competition—threats posed by other members of one’s own species.
This latter category further breaks down into inter-group (e.g. raiding, warfare) and
intra-group (i.e. inter-individual) forms.
Of these many varieties of competition, only one subset—biotic, intra-specific,
intra-group competition—pits the individual against other individuals in their group.
In every other instance, the individual’s own survival is positively correlated with that
of the other members of their group, since their ability to overcome such challenges is
highly dependent upon the presence and cooperation of their fellow group members.
All those extra eyes and ears increase foraging efficiency and vigilance, while all
those extra arms and legs augment the individual’s capacity for self-defense in the
face of both inter- and intra-specific challenges.
Most importantly, proximity to, and the goodwill of, others provides the individual
with a priceless buffer against the vagaries of life, since said others can provision the
individual with food, water, transport, and protection during those inevitable periods
when injury, illness, childbirth, or other temporary disability means they can’t obtain
these for themselves. Without comrades able and willing to do this, such episodes
easily turn lethal.
54 D. A. Marshall
The ultimate rejoinder to the idea of human maximizers is the familiar phe-
nomenon of satiation. However cold, thirsty, tired, or hungry an individual may
be, once sufficient warmth, water, sleep, or food is obtained, further consumption
becomes aversive. That all of humans’ homeostatic drives are similarly subject to
satiation is compelling evidence that we are not only not naturally maximizing, but
innately disposed against it.
Recognizing the temporal dimension of nutrition also brings the true nature of
the challenge, and yet another essential distinction between rationalist and adaptive
logics, into focus: Though an organism may be unable to make use of any more
calories at time “A”, they will inevitably need more at some not-too-distant time
“B”. Even in the narrow domain of nutrition, the essential problem is not one of
maximization, but the substantially different and more complex one of continuity.
The same applies to every other dimension of survival: Being warm or protected
today does nothing to negate one’s odds of freezing to death or being ambushed
tomorrow. In conjunction with the aforementioned lethality of single-mindedness,
this suggests that human behavior has evolved not to emphasize having as much as
possible of anything, but rather to always have enough of everything. In a nutshell,
survival is a lot more like juggling than throwing shot-put.
Informed. The idea that humans make decisions in light of complete informa-
tion has always been a straw argument, since it is obvious that crucial data is often
unavailable, or exceeds human perceptual and memory capacities. But the bigger
obstacle to a fully-informed individual actor is less one of ability than of motiva-
tion. Ironically, it is the capacity for human rationality which insures that, left to
our own devices, we seldom seek to acquire anything like “complete information”.
Understanding why requires a bit of background:
Across the natural world, most forms of behavior are effectively under control
of the environment (i.e. correlated with daylength, temperature, etc.). But among
humans, the evolutionary premium on ever-faster responses to changing conditions
produced a capacity for flexible, internally instigated and directed behavioral control
[7]. This is possible thanks to our capacity for constructing an internal model of the
world—a “schema” or “nomos” [8] that makes the internal simulation of potential
behaviors and their consequences—in a word, calculation—possible.
The individual’s dependence upon that internal model creates competing epis-
temic needs that regularly take priority over those for accurate information, as actors
balance accuracy against the need to protect their nomos from epistemic threat. Put
simply, we’d rather believe we are right than be right, so individuals are typically
highly selective in their search for, and weighting of, information, so as to avoid or
downplay nomos-inconsistent data.
More generally, while rationality puts a premium on complete and accurate infor-
mation, biological adaptation is much less persnickety, in that incomplete and inac-
curate is often good enough for its purposes—and sometimes even superior. For
example, as Harari points out, large-scale human sociality has long depended upon
individuals’ ability to subscribe to such patently imaginary constructs as nation and
ideology [9]—the implication being that our success as a species is a function of our
generally credulous and non-critical stance towards information.
56 D. A. Marshall
Fitness—an organism’s ability to leave viable copies of its genotype—is the undis-
puted currency of evolution. Having elaborated the ways in which survival is ill-
served by rationality, the task here is to demonstrate that even if rationality were
an asset to survival, it still wouldn’t be an evolutionary imperative, since fitness
entails much more than survival—and its other elements are no better fulfilled by
instrumental rationality than is survival.
So just what is fitness? One venerable answer is that it is the ability to have
offspring that can have offspring, a formulation which still well-captures the breadth
and depth of the evolutionary challenge facing organisms. Thus, for most intents and
purposes, fitness comes down to an organism’s ability to make grandchildren.
How to Make Grandchildren. Making grandchildren, and thereby achieving
fitness, is a complex and multi-dimensional task, But the problem is parsable into a
handful of sub-tasks. In roughly chronological order, these are:
Task 1—Survive to (and Through) Sexual Maturity. Though survival is not the
entire challenge of fitness, it is something of a prerequisite. Given its elaboration
above, let us simply reiterate the point that, contrary to rationalist models, it is
fundamentally an exercise in continuous, multi-dimensional satisfycing.
Task 2—Gain and Maintain Access to Potential Reproductive Partners. The issue
of reproductive access has both quantitative and qualitative dimensions. Quantita-
tively, the relatively low probability of conception per act of intercourse among
humans implies a fairly linear relationship between an individual’s frequency and
duration of access and their fecundity. Qualitatively, there is a premium on partners
in possession of particularly fitness-conveying genes and/or socio-cultural capital.
Though males’ interests arguably lie more on the quantitative side while females’
interests arguably lie more in the qualitative realm, both have an interest in gaining
and maintaining access to the largest possible pool of potential partners, as it makes
both a greater number of, and a greater selectivity among, partners possible.
The Biological Logic of Human Action: On the (Considerable) … 57
Having worked our way through three discrete ways in which the assumptions of
evolved rationality are untenable, we now work our way back through these argu-
ments in order to develop an alternative, theoretically and empirically credible logic
of human action.
58 D. A. Marshall
In terms of biotic threats, the group’s redundancy and wide dispersal of eyes,
ears, and attention significantly decreases each individual’s vulnerability to macro-
predators and conspecific attackers by making the detection of such threats more
likely [12]. For example, though circadian rhythms mean that each individual organ-
ism is asleep, and thus singularly vulnerable, during some part of each day, inter-
individual differences in sleep patterns within any group of conspecifics mean that
the group as a whole has fewer and smaller lapses of vigilance, and thus a reduced
vulnerability to surprise attack relative to any individual. Moreover, once the group
detects a threat, it can address it more effectively by mounting a collective defense or
counterattack. This capacity for collective counter-aggression also deters the kind of
conspecific inter-group violence that characterized premodern human life [14]. Note,
too, that each group’s ability to deter attack and defend itself is directly proportional
to its size, its solidarity, and the health of its individual members.
We come at last to the inter-individual competition which the rationalist model
assumes to be ubiquitous. Here, one’s fitness really might come at the expense of
another’s, potentially evoking selective pressures for individualist behaviors. Without
denying that such competition exists, the pressure it exerts is insufficient to instill
individualism as a defining feature of human nature because it is, as noted, but one
threat to survival among many- the rest of which collectively present formidable
selection pressures against it. Moreover, even inter-individual competition has a
cooperative dimension, in that success in such instances is more often decided by
the size and status of one’s allies than by one’s individual strength.
Task 2—Reproduction. Group living is obviously advantageous when it comes to
maintaining access to a sizeable pool of potential partners. Among humans, with our
frequent and recurrent periods of concealed fertility, near-continuous sexual receptiv-
ity, and frequent recourse to polygamy/polygyny, more-or-less constant proximity
to potential partners is integral to fitness. Beyond the mere quantity of potential
partners, group living provides a context in which to evaluate their potential contri-
butions to offspring, since some fitness-enhancing traits—e.g. charisma, status, intel-
ligence—become apparent only upon the kind of extended observation that group
living allows.
Task 3—Mother-Child Viability. Sociality is most decisive when it comes to
insuring that the mother-child pair survives their pre-natal, peri-natal and post-natal
interdependency. Pregnancy, childbirth, and its immediate aftermath represent an
extended instance of temporary debility for the mother, during which other group
members become indispensable provisioners of their survival needs. Following birth,
the child’s growing nutritional needs soon outstrip any individual’s ability to pro-
vide them by herself, so the alloparenting (communal nursing, kiss-feeding) that the
group provides proves indispensable to the pair’s survival, and thereby to her and the
father’s, fitness [15].
Task 4—Offspring Survival. Any offspring’s ability to survive from weaning
to sexual maturity is complicated by the fact that humans are born “unfinished”,
bereft all but a token arsenal of instincts [8]. The period during which offspring are
acquiring their nomos (i.e. “childhood”) is one of ignorance-fueled vulnerability. A
child’s odds of surviving are, yet again, very much a function of their membership
60 D. A. Marshall
in an intact and formidable group that can protect them from said ignorance. Here,
too, both parents’ fitness is dependent upon; (a) the existence of such a group and,
(b) their offspring’s ability to join and remain a member of that group.
Task 5—Socializing Offspring. Because human beings are born “unfinished”,
both parents’ fitness is tied to their offspring having acquired the knowledge and
practical skills necessary to their own success at all the tasks of fitness in their
particular physical, social, and cultural contexts. This requires the kind of reiteration
and repeated reinforcement, ubiqutous behavioral modeling, and extensive practice
that no parent, or parents, can provide by themselves.
In sum, it really does take a village to make grandchildren—and thereby for
individuals to achieve evolutionary fitness. But villages don’t occur by chance—their
emergence is dependent upon the existence and ubiquity of individuals possessing
an extensive set of behavioral capacities and dispositions that make them possible.
For such traits to be so invariant among human beings that tribes and societies can
reliably be assembled from them suggests the influence of strong and consistent
selection pressures throughout human evolution.
The lesson of “Biology’s Blunt Tool” in Part I is that however adaptive any particular
trait might be, there is no guarantee that it actually will evolve. In the abstract, this is
as true of the traits of human sociality as of human rationality. But human sociality
was much more likely to evolve than human rationality for at least three reasons.
In the first place, sociality is a bigger target than is instrumental rationality. The
latter involves the co-occurrence of several very specific capacities and dispositions,
the probability of each of which is small, and the cumulative probability of all is
vanishingly so. Sociality is a larger target in that there are many different ways of
being social—any of which are consistent with the hypothesis that some kind of
sociality would evolve. While the particular ways that humans are social very much
matters, the biological model doesn’t try to deduce the particulars ahead of time, but
instead induces them from the empirical literature.
Secondly, relative to rationality, sociality has a much longer runway. Since only
humans have typically been ascribed the quality of rationality per se, one presumes
that its constituent traits would have had to evolve within the three hundred-thousand
year history of homo sapiens, or at most, the three million year history of the homo
genus. Either figure is an implausibly short time for the several defining features of
instrumental rationality to have arisen, been selected for, and dispersed throughout
the species. By contrast, the pedigree of sociality stretches back hundreds of millions
of years. The net effect was that our species was evolutionarily primed for sociality,
making further social adaptations more likely to emerge over time because the genetic
pathways to them were already well-established [12].
The Biological Logic of Human Action: On the (Considerable) … 61
Finally, sociality has a clear leverage advantage over rationality. As we’ve seen,
sociality is advantageous in multiple ways across each of the tasks of fitness, increas-
ing the number of ways and reasons a prosocial trait can be selected for once it
emerges. Furthermore, it benefits from a positive feedback loop, in that the advent of
sociality brings new, socially-mediated selective forces like reputation and assortive
association into play.
But sociality has an even bigger lever working for it, in the form of inter-level selec-
tion. Here—in the conflict between a lone individual and a collective competitor—is
where the real fulcrum of sociality lies. In a population of competing individual
organisms, any two individuals who can mutually suppress their competition against
one another enough to cooperate, even minimally, against their solo competitors has
a decisive advantage over them in every dimension of fitness, rendering individuality
a clear and ongoing liability.
Having established that the evolved biological logic of human action is different from,
and more often than not antithetical to, its rationalist predecessor, let us consider a
handful of the implications thereof, as regards individual experience, social science,
social policy, and government.
The argument herein has not been that we, as humans, are incapable of instrumental
rationality, only that it is not a biologically evolved, hence innate and universal,
characteristic of human beings. Obviously, we possess some nominal capacity for it,
since we would otherwise have no means recognizing the ways and degrees to which
we deviate from it. Moreover, though it did not evolve per se, our (limited) capacity
for rationality is nevertheless a by-product (or spandrel) of other capacities that did
evolve—and it certainly has been a significant asset to human fitness in some ways
at some times. But to argue that it constitutes the underlying logic of human behavior
would be tantamount to calling our species homo accordionus based on the fact that
some of us, sometimes, can play the accordion.
This (limited) capacity for instrumental rationality is a recent addendum to a much
older and more robust biologically adaptive system. Much of the tragedy and grandeur
of human existence ultimately comes down to the fact that we are homo duplex [21]—
composed of and subject to the competing demands of two different control systems:
The evolved, non-conscious, and automatic “System 1” (i.e. the biological logic
established here), and; The newer, cultivated, conscious, and controlled “System
2” [4, 22]. The tension between these is loaded with implications for human life
and history [23, 24], but for present purposes, we will confine our attention to the
simple observation that rationality is, like flight (and accordion playing) not a human
property, but merely a human potential, one which comes neither naturally nor easily
to us. The older System 1 is primary to System 2 in every sense of the term. It is the
default logic of human action, overcomeable only with much effort and attention.
The obvious question of where and when individuals manage to realize what degree
of rationality brings us to a second implication.
64 D. A. Marshall
For all of its implausibility as the logic of human action, instrumental rationality
remains an important concept in social science—as a Weberian ideal-type against
which to contrast observed behavior, as a normative assumption of Western cul-
ture, and, again as per Weber, as a (the?) defining feature of modern society. Most
importantly, though instrumental rationality fails as an independent variable—an
assumption with which to describe or explain human behavior—it is a promising
dependent variable—an outcome to be explained as a function of the problem at
hand, of individual motivation, and of the cultural and social contexts in which it
occurs [25].
As per the “massive modularity" (i.e. “swiss army brain”) hypothesis, the human
mind is an assembly of discrete special-purpose calculators, rather than a single
general-purpose processor [26], so our ability to deal with a problem rationally
depends upon its falling into a domain that both lacks competing non-rational heuris-
tics, and for which our calculation skills are evolutionarily primed (e.g. deception
detection [26]). Additionally, the odds of a problem evoking rationality are increased
when it is highly circumscribed, such that the option set is limited and explicit,
the relevant probabilities are pre-established, and the relevant information is highly
domain-specific and readily available. At the same time, some individuals are more
likely to engage in the effort necessary to move beyond the ready answers provided
by System 1 than are others as a result of personality, experience with “critical think-
ing”, and geographical context (i.e. rural fixity vs. urban mobility), among other
factors [27].
The potential for rationality is also a function of the cultural context of the problem,
in terms of the “tools” it does or doesn’t provide to support rational decision-making,
from stopwatches to software packages, double-entry bookkeeping, and t-tests. Less
obvious social technologies—i.e. the scientific method and peer-review—exploit
low-moral-density (loosely-coupled) social structures to dilute and mitigate many
of the individual obstacles to rationality. As for the social context of a problem,
though instrumental rationality was maladaptive for most of the history of human
evolution, it is not necessarily so today. Many of the adaptive challenges that once
made it so antithetical to fitness are much less relevant in the modern world, where
the capitalist economy has shifted more of human ambitions into the fungible and
insatiable domain of money. Instrumental rationality is potentially less maladaptive
and more profitable than it used to be, which sets the stage for a new domain of (now
cultural) evolution—one effectively decoupled from biological fitness and biological
evolution and focused instead on the acquisition of resources.
there are good reasons to at least supplement the idea of the individual social actor
with what is arguably the more naturally-occurring unit of human action—the tribe.
If anything, such tightly-coupled groups tend to amplify individual human traits
and exacerbate non-rationalist tendencies by compounding individual biases and
dispositions. While vestiges of such tribes still abound (e.g. families, squads, cults),
they are eclipsed in the modern era by a third kind of social actor—the loosely-
coupled organization. Unlike the other two actors, the organization is distinguished
by its instrumentally rational nature.
Under the constraints of an emergent imperative to maximize their acquisition of
material resources, organizations have (culturally) evolved to leverage the finite ratio-
nality of their constituent persons into potentially infinite manifestations of zweck-
rationality. Lacking a physiology of its own, the organization is free from many of the
competing needs and limitations that beset individuals. It imposes an incentive struc-
ture that aligns different individual goals with its own, and parses decision-making
into myriad highly specialized and routinized domains, thereby concentrating the
meager potential for rationality of its individual members to produce instrumental
rationality at scale. Even as it concentrates cognitive power, it, by a combination of
happenstance and design, dilutes human morality by diffusing individuals’ sense of
responsibility for their actions both horizontally and vertically [25].
The emergent rationality of organizations explains how instrumental rationality
came to be the master trend of the modern world even as the individual human beings
who are its original constituents remain largely unable and unwilling to engage in
it. It also alerts us to a potentially era-defining social problematique of the modern
world—the uneven distribution of rationality. In a world in which the biological
logic no longer serves fitness much of the time, and in which profit is the prevailing
maximand, the old, naturally-occurring social actors of individuals and tribes are
massively outmatched by their new, organizational, competitor. Witness the extent
to which organizations are presently accruing unprecedented economic, political,
and cultural power at the expense of individuals and tribes, in the form of widening
inequalities, subzero tax rates for large companies, and corporations’ increasing
arrogation of the rights and privileges of human citizens.
One of the most disturbing mysteries of the early twenty-first century has been the
failure of democracy to secure benefits for voters because those voters repeatedly
vote against their own interest. As per one popular treatment, “What’s the matter with
Kansas?” [28]. From the perspective of this chapter, this pattern is fully predictable.
Democracy, as noted in the introduction, is predicated on the existence of a more
or less rational electorate—one that is, at minimum, self-interested enough to seek
accurate and complete information, and to engage in the calculation necessary to
choose the best way to pursue those interests. But as we’ve seen, human beings have
never truly fit that description. Following their advent and ascent, organizational
66 D. A. Marshall
actors, consciously or not, discovered that the biological logic of human action can
be deployed against the individuals who make up an electorate, and over time, they
have perfected the technologies of doing so. Though voters may think of themselves
as informed, calculating, and self-interested, they are easily manipulated into acting
against their own interest by a combination of fearmongering and manufactured
tribalism which plays upon their needs for belonging, their fear of ostracism, and
their need for collective identity (e.g. racism, nationalism).
That this strategy is only now coming to fruition has much to do with the rise of
technologies which made possible, for the first time, the mass construction of the
kind of fully-insulated epistemic bubbles that truly effective propaganda requires
[29]. Thus, we close this chapter with an illustration of one of the dangers inherent
in the rationalist assumption: The internet was supposed to usher in the next phase
of humanity by democratizing access to information and calculation power—but
such rosy assumptions were predicated on false conception of individuals as rational
consumers of information and calculating maximizers. We were thus blindsided by
the reality that human actors were never going to use it that way, opening the way
for organizational actors—from Fox news to “Fancy Bear” to Facebook—to exploit
human nature in order to seduce voters into behaving in ways that are profoundly
antithetical to their own self-interests by using their biological logic against them.
References
1. Weber, M.: Economy and Society. University of California Press, Berkeley CA (1978 [1922])
2. Fiske, S., Taylor, S.: Social Cognition, 2nd edn. McGraw-Hill, New York NY (1991)
3. Arieley, D.: Predictably Irrational: The Hidden Forces that Shape Our Decisions. Harper, New
York NY (2009)
4. Gould, S.J., Lewontin, R.: The spandrels of San Marco and the panglossian paradigm: a critique
of the adaptationist programme. Proc. R. Soc. Lond. Ser B Biol. Sci. 205, 581–598 (1979)
5. Darwin, C.: The Origin of Species: 150th Anniversary Edition. Signet Reprint (2003 [1859])
6. Kahneman, D.: Thinking, Fast and Slow. Farrar, Strauss, Giroux, New York, NY (2011)
7. Bonner, J.: The Evolution of Culture in Animals. Princeton University Press, Princeton, NJ
(1980)
8. Berger, P., Luckmann, T.: The Social Construction of Reality: A Treatise in the Sociology of
Knowledge. Anchor Books, Garden City, NY (1967)
9. Harari, N.Y.: Sapiens: A Brief History of Humankind. Harper Perennial, New York, NY (2018)
10. Bargh, J., Chartrand, T.: The unbearable automaticity of being. Am. Psychol. 54, 462–479
(1999)
11. Nowak, M.: Supercooperators: Altruism, Evolution, and Why We Need Each Other to Succeed.
Free Press, New York, NY (2011)
12. Wilson, E.O.: Genesis: The Deep Origins of Societies. Liveright, New York, NY (2019)
13. Kaplan, H., Hill, K., Hawkes, K., Hurtado, A.: Food sharing among Aché hunter-gatherers of
Eastern Paraguay. Curr. Anthropol. 25, 113–115 (1984)
14. Bowles, S., Gintis, H.: A Cooperative Species: Human Reciprocity and Its Evolution. Princeton
University Press, Princeton, NJ (2011)
15. Hrdy, S.B.: Mother Nature: Maternal Instincts and How They Shape the Human Species.
Ballantine Books, New York, NY (2000)
The Biological Logic of Human Action: On the (Considerable) … 67
16. Diener, E., Lusk, R., DeFour, D., Flax, R.: Deindividuation: effects of group size, density,
number of observers, and group member similarity on self-consciousness and disinhibited
behavior. J. Pers. Soc. Psychol. 39, 449–459 (1980)
17. Tajfel, H.: Experiments in intergroup discrimination. Sci. Am. 223(5), 96–102 (1970)
18. Lieberman, M.: Social: Why Our Brains are Wired to Connect. Crown Publishers, New York,
NY (2014)
19. DeWaal, F.: Primates and Philosophers: How Morality Evolved. Princeton University Press,
Princeton, NJ (2006)
20. Weisner, P.: Taking the risk out of risky transactions: a forager’s dilemma. In: Salter, F. (ed.)
Risky Transactions. Berghahn Books, Oxford UK (2002)
21. Durkheim, E. The dualism of human nature and its social conditions. In Wolff K (trans.) Emile
Durkheim: Essays on Sociology and Philosophy, pp. 325–340 (1964 [1914])
22. Chaiken, S., Trope, Y.: Dual Process Theories in Social Psychology. Guilford Press, New York,
NY (1999)
23. Marshall, D.A.: Behavior, belonging, and belief: a theory of ritual practice. Sociol. Theor.
20(3), 360–380 (2002)
24. Marshall, D.A.: Sacralization and the throes of modernization: witch trials, world war, and the
WTC. New York J. Sociol. 1, 26–90 (2008). www.newyorksociology.org
25. Marshall, D.A.: Beyond a Rational Choice Sociology: A Sociology of Rationality. Doctoral
Dissertation. Retrieved from Proquest Dissertation Publishing (2003)
26. Barkow, J., Cosmides, L., Tooby, J.: The Adapted Mind: Evolutionary Psychology and the
Generation of Culture. Oxford University Press, New York NY (1992)
27. Stanovich, K.: Who is rational?: Studies of Individual Differences in Reasoning. Psychology
Press, London UK (1999)
28. Frank, T.: What’s the Matter With Kansas?. Henry Holt, New York, NY (2004)
29. Ellul, J.: Propaganda: The Formation of Mens’ Attitudes. Vintage Books, New York, NY (1973)
Social Media: “Surrogate Tribes”?
Paolo R. Crocchiolo
1 Introduction
P. R. Crocchiolo (B)
American University, Rome, Italy
e-mail: paolo.crocchiolo@libero.it
evolved as a complex machinery constantly monitoring our inner state. The “emo-
tional colouring”, which is generated by the NHS, and characterizes all activities
performed by our mind, thus appears to be indispensable for the preservation of the
vital processes generating that “well-being” state.
Any “species”, including ours, can be viewed as a “DNA consortium”, an involun-
tary and unconscious way of survival and reproduction that allowed the propagation,
so far, of their respective genetic pools. We, “individuals”, aren’t, in this perspective,
but physically separated variants of the shared broader DNA consortium, i.e. of our
common genetic pool.
Within this context, individuals were selectively advantaged who best fit the needs
of their troop/tribe’s social structure (in its turn emerged from a previous, in fact a
parallel, selection), as they were carrying the tribal instincts that best granted their
community’s genetic pool’s survival and reproduction [3]. Thus, individual selection
compounds in the long run with social selection.
In this context, social media, and networks, may be viewed as a sort of “surrogate
tribes”, i.e., updated instruments of a contemporary version of the evolved human
social communities.
According to evolutionary biology, our very complex individual phenotypes are
the end-results of an equally sophisticated gene-network, i.e., of genes unconsciously
interacting with each other in such a way that their end-product made them survive
in preference to the end-product of other networks/patterns, the latter being outnum-
bered because of the former’s higher reproduction rate in their given natural and
social environment.
In this context, both our body and our mind seem to be the result of biological
evolution by natural selection [4, 5]. More specifically, both physical and psycholog-
ical traits of our phenotypes appear to represent the end-results of a complex chain
of processes triggered by our genotypes’ proteinopoietic activity.
In fact, our body which, as a sort of “envelop”, harbours and protects our genes,
and our mind, whose neural correlate is our complex neural-hormonal system (NHS),
appear to have evolved (by selective differential reproduction rate of their phenotypic
carriers) in function of protecting our body’s integrity, thus ultimately enhancing our
genes’ reproductive potential [2].
In this perspective, mind-body units may be interpreted as “survival and
reproduction machines” of their underlying genes [6].
Representing the biological interfaces with the outer world, the senses appear to
have evolved as prostheses of our neural systems. Senses, in fact, may be defined
Social Media: “Surrogate Tribes”? 71
One of the main emergent properties of our NHS is its fundamental capacity to
distinguish between positive and negative hedonistic tones (pleasure, pain, and every
sensation in between the two ends of the perceivable emotional spectrum) [1].
The latter capacity rests upon an elaborate network connecting certain inner brain
areas, mainly located in the “limbic system”, with specialized groups of nervous cells
secreting neurotransmitters, such as dopamine, serotonin, etc., and neuro-hormones,
such as endorphins, oxytocin, etc. All these molecules, when released upon recogni-
tion of an appropriate triggering signal by the cells where they are normally stored,
travel to, and impact on, their specific receptors situated on their corresponding target
cells.
Natural selection rewarded those phenotypes endowed with a NHS that associated
a positive or a negative hedonistic tone, respectively, with situations and events
enhancing or inhibiting their survival and reproduction, thus originating reflexes
and instincts. The latter may be defined as genetically transmissible associations
of neural-hormonal rewarding/penalizing mechanisms with their resulting specific
survival- and reproduction-enhancing skills/potentials. Instincts, in other words, may
be viewed as a kind of “unconscious bridles” that guide us in our thoughts and in
our actions.
72 P. R. Crocchiolo
In addition to reflexes and instincts, the hedonistic tone system also originates
emotions and feelings.
Emotions may be defined as amplifying modulations [1] of the basic, positive
or negative, DNA-imprinted hedonistic tones, allowing their carrier organisms to
specifically adapt their behavioural responses to the great variety of natural and
social environmental challenges they are exposed to. Feeling, in man, represents a
further fine-tuning capacity resting upon the development of autobiographic self-
consciousness [2].
Why modulations? Qualitatively, organisms developed the capacity to select,
within a generally positive or negative hedonistic background, a number of different
neural-hormonal patterns appropriately fitting different situations to cope with (e.g.,
fear, jealousy or rage predisposing to different behaviours within a generally negative
hedonistic tone).
And why amplifying? Quantitatively, organisms developed the capacity to tailor
the intensity of their emotions to the biological impact of the triggering events.
Amplification of the emotional colouring roughly reflects the increase, or decrease,
in biological fitness of the carrier’s organism in relation to the initial situation.
The broader the range of possible modulations and the ampler the intensity of the
positive or negative emotion evoked, the higher the likelihood of the underlying gene
combinations to survive, reproduce, and spread within the subsequent generations.
While reflexes and instincts are phylogenetically selected and consolidated by adap-
tation to a relatively constant environment, emotions and feelings are ontogenetically
learned and potentially adaptable to individual experience. Neural-hormonal mech-
anisms, such as the “mirror neurons” system [3], grant the necessary flexibility to
adapt the organism’s emotional response to a virtually infinite variety of triggering
events and situations it may be exposed to during the course of its life. Thus, organ-
isms gradually acquired the ability to qualitatively and quantitatively fine-tune the
appropriate responses tailoring them to the specific situations encountered, choos-
ing from a more basic repertoire of reflexes/instincts directly tied to survival and
reproduction, and made available by ancestral, previously consolidated structures
and functions [2].
In this respect, memory plays an analogous role: qualitative and quantitative fine-
tuning of the response is related to a learning process whereby previous individual
experiences exhibit a booster effect, biasing the organism’s selection and modulation
of the most appropriate emotional response, based on the previously retained/stored
experience.
Social Media: “Surrogate Tribes”? 73
Thus, natural selection rewarded, and spread through the next generations, those
genetic blends that codified neural-hormonal patterns of response to environmental
challenges which induced behaviours best granting, in the average, the reproductive
success of their phenotypic carriers [8].
As the NHS is not a static phenomenon, but a dynamic process based on a continuous
turnover, it changes with time. In fact, our gene-networks’ activity unfolds in a
complex feedback-regulated time sequence.
According to evolutionary psychologists [4], neuro-embryological development
and, subsequently, neural-hormonal phases of development characterizing the course
of our life after birth, seem to be genetically imprinted and pre-programmed as a result
74 P. R. Crocchiolo
Within this context, but at a subordinate level, those individuals were selectively
advantaged, who best fit the needs of the troop/tribe’s social structures, which in the
meantime had emerged from a previous, in fact from a parallel, selection. According
to sociobiologist Edward O. Wilson [12], troop- and tribal identities, i.e., the sense
of belonging to their community, as well as troop- and tribal hierarchies, were then
selected among our ancestors as survival- and reproduction enhancing instinctual
adaptations, to the extent they were protecting the integrity of the collective “en-
velop”. The selected individuals were the ones carrying the tribal instincts that best
granted the success of their respective community’s acquisition, and defence from
other tribes and predators, of the food and space necessary for their genetic pool’s
survival and reproduction.
Genes were selected that predisposed their phenotypic envelops to the sharing of
tribal behaviours, beliefs, practices and emotional involvements, including collective
excitements such as playing, dancing, singing, etc. These instinctual behaviours all
contributed to inclusion of their individual organisms into their surrounding, mainly
next of kin, community [13]. Adherence to the shared “biased thinking” (tribal prej-
udices, taboos, commonly accepted stereotypes, etc.) turned out to be advantageous
for their carriers, hence psychologically rewarding (Facebook’s “I like it” may sound
like a pale vestige of one of our ancestral reward-system arousal factors). Isolation,
self-exclusion, and marginalization instead, all resulted in selective disadvantages,
Social Media: “Surrogate Tribes”? 75
leaving less chances to survive and reproduce to the carriers of those alternative
exclusion-inducing instinctive traits, and hence of their underlying genes.
3.3 Language
We have the potential to use to their full extent the instruments made available to
us by our NHS-endowed “survival and reproduction machines”. Among all these
instruments, language plays a paramount role to communicate, both rationally and
emotionally, within human societies. In fact, since perceptions are automatically and
rapidly translated into language, we have evolved as verbal creatures [14].
In the sort of “ping-pong” game called sexual selection, gene combinations were
selected that—via NHS—“suggested” attraction to certain physical or psychological
traits of their potential partners’ phenotypes. Peacocks, for instance, that instinctively
exhibited their magnificent tails (a powerful proxy standing for “good genes”), were
involved in the reproductive success of peahens that in turn had been selected to the
extent they were impressed and attracted by those very tails, which ultimately had
granted a higher reproduction rate to their peacock carriers in the first place.
Through the media, in general, and the social media in particular, a possibility is
offered to the “human peacocks, and peahens” to exhibit, as a kind of “ornament”,
their very “intelligence” [15], or other mental or physical attributes, to their potential
partners in the web. The social media thus convert, also from this point of view, into
convenient instruments to convey all kinds of messages (not only “friendship”) to
the other members of the modern tribal community.
76 P. R. Crocchiolo
4 Conclusion
In the long history of evolution, the genes that exhibited best survival and repro-
duction chances were those that, in addition to mono- or multi-cellular “individu-
al” skills, evolved particularly strong “social bonds”. In fact, the same cooperative
interactions observable in all living beings, starting from bacterial strains, social
insects, flocks of birds, mammals’ troops, etc., seem to be amenable to the same
gene-preserving driving force that generated the social bonds characterizing human
tribes, subsequently evolved into today’s complex human societies. The gradual evo-
lution of more and more sophisticated “emotional brains” [8] in multicellular animal
organisms, obviously played a fundamental role in this context.
Thus, in conclusion, social media may be seen as updated instruments of a contem-
porary version of the human social groups, a sort of “surrogate tribes”, instinctively
dictated, in ultimate analysis, by our genes.
References
1. Johnston, V.S.: Why We Feel: The Science of Human Emotions. Perseus Books, Reading, MA
(1999)
2. Damasio, A.R.: The Feeling of What Happens: Body and Emotion in the Making of
Consciousness. Harcourt Brace, New York (1999)
3. Rizzolatti, G.: Mirrors in the Brain: How Our Minds Share Actions, Emotions, and Experience.
Oxford University Press, Oxford (2008)
4. Cosmides, L., Tooby, J.: Cognitive Adaptations for Social Exchange. The Adapted Mind:
Evolutionary Psychology and the Generation of Culture. Oxford University Press, New York
(1992)
5. Damasio, A.R.: Self Comes to Mind: Constructing the Conscious Brain. Pantheon Books, New
York (2010)
6. Dawkins, R.: The Selfish Gene, 3rd edn. Oxford University Press, Oxford (2006)
7. Boncinelli, E.: Minuscoli universi chiamati “uomini”. Corriere Della Sera, Milan, 28 Nov
(2003)
8. Le Doux, J.E.: The Emotional Brain: The Mysterious Underpinnings of Emotional Life. Simon
& Schuster, New York (1998)
9. Bernard, C.: An Introduction to the Study of Experimental Medicine. Dover Publications, New
York (1957)
10. Metzinger, T.: Being No One: The Self-model Theory of Subjectivity. MIT Press, Cambridge,
MA (2003)
11. Gould, J.S.: The Structure of Evolutionary Theory. Belknap Press of Harvard University Press,
Cambridge, MA (2002)
12. Wilson, E.O.: Consilience: The Unity of Knowledge. Knopf, New York (1998)
13. Ridley, M.: Nature Via Nurture: Genes, Experience, and What Makes Us Human. Harper-
Collins, New York (2003)
14. Pinker, S.: The Language Instinct: How the Mind Creates Language. HarperCollins, New York
(1994)
15. Miller, G.: The Mating Mind: How Sexual Choice Shaped Human Nature. Vintage, New York
(2001)
Moral Bubble Effect
Violent and Unaware of It: How Difficulties
in Recognizing One’s Own Violence Lead to Disregard
the Inflicted Harm
Lorenzo Magnani
In 1999 Justin Kruger and David Dunning published in the Journal of Personality and
Social Psychology the seminal paper “Unskilled and Unaware of It: How Difficulties
in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments” [10],
which explained what later on has been called Dunning-Kruger effect. In the present
article I would like to offer to the attention of sociological, psychological, and psychi-
atric research a similar effect that regards the absent (or at least extremely unstable)
awareness of human violent acts, as indicated in the subtitle “Violent and Unaware
of It: How Difficulties in Recognizing One’s Own Violence Lead to Disregard the
Inflicted Harm”. I have already sketched this effect in my book Understanding Vio-
lence. The Intertwining of Morality, Religion, and Violence: A Philosophical Stance
[14, chapter three]. I am detailing here some more hypotheses that ask for further
L. Magnani (B)
Department of Humanities, Philosophy Section and Computational Philosophy Laboratory,
University of Pavia, Pavia, Italy
e-mail: lmagnani@unipv.it
1 Moral rules and related violence are first of all performed thanks to natural language—imagine
for instance the military/violent status of hateful, racist, homophobic speech—but also by motor
actions and emotions: it is well-known that patent hostility in emotions is a chance to start violent
actions.
2 Studying the related problem of “feeling of knowing” Burton [4, p. 12] contends that it is in general
recognized by the absence of knowledge itself (that is when we do not know something we feel
we should know), and it is also connected to the well-known situation of the so-called “cognitive
dissonance”: “In 1957, Stanford professor of social psychology Leon Festinger introduced the term
cognitive dissonance to describe the distressing mental state in which people find themselves doing
things that don’t fit with what they know, or having opinions that do not fit with other opinions they
hold”.
Moral Bubble Effect 79
As before, when in an epistemic bubble, cognitive agents always resolve the tension between
only the apparently true and the genuinely true in favour of the genuinely true. Corollary
12a reminds us of the remarkable perceptiveness of Peirce […] that the production of belief
is the sole function of thought. What Peirce underlines is that enquiry stops when belief is
attained and is wholly natural that this should be so. However, as Peirce was also aware,
the propensity of belief to suspend thinking constitutes a significant economic advantage. It
discourages the over-use of (often) scarce resources [22].
Hence, truth is “fugitive” and the process of embubblement stresses the importance
of corrigibility of our beliefs, especially the ones that present unwelcome effects
during the processes of problems solving and decision-making: Woods correctly
contends that the process of embubblement is not corrigible in itself, unless the
agent dismisses a bubble to adopt a new one. The issue about corrigibility seems to
have something in common with another issue, namely, de-biasing. Indeed epistemic
bubbles share with biases the fact that both present a resistance to correction.3
3 Even though, during the last decades, studies regarding decision-making have increased knowledge
about biases and fallacies, de-biasing inferential processes are still relatively uninvestigated. Some
more notes about this issue are illustrated in [14, chapter three].
80 L. Magnani
and eradicate her own misconceptions. Thanks to the autoimmune mechanism the
agent can enter a status that is less problematic as possible, for example in case of
decision-making procedures and emotional reactions. The intertwining between the
epistemological status of the agent and her related cognitive and emotional state is
indeed at the basis of the concept of cognitive autoimmunity. 4
The studies on logic, informal logic, and rhetoric have demonstrated that fallacies
performed thanks to human language are often characterized by their “military” (in
Thomian sense) effects in terms of softness and gentleness, because one of their main
virtues is the capacity to disguise errors. Epistemic bubbles are complicit in these
processes of concealment, and being unaware, in a constitutive and spontaneous
way, of our mistakes is in many cases intertwined with the self-conviction that we
are not at all violent and aggressive in the arguments we are advancing (and in
our possible subsequent acts). We have to remember that humans use language and
the so-called fallacies embedded in it to the aim of producing positive and vital
results5 —even if they can at the same time carry violent outcomes: very often the
eventual target agent feels violently affected by fallacious expressions. If fallacies
are, so to speak, eco-cognitively fruitful, it is difficult to call them again “fallacies”,
at least if we use the term fallacy in its standard negative connotation. I contend that
the example of the role of fallacies embedded in human argumentation favors the
hypostatization of the more general6 phenomenon of what I called moral bubbles
[15], which are homomorphic with the epistemic bubbles. In summary: unawareness
of our mistakes is often intertwined with the absence of awareness concerning the
deceptive/aggressive status of our speech (and behavior).
4 We have to remember that the cognitive autoimmunity of the agent is also related to the psycho-
logical problems of the so-called “epistemic feelings” [17, 18], that is feelings that are linked to the
epistemic situation of the agent, such as in the case of feeling of knowing, feeling of forgetting, and
tip on the tongue feeling. They can be seen as the neurological and cognitive detonators of cognitive
autoimmunity [2].
5 Just to make an example, research on informal logic has illustrated that a fallacy such as hasty
generalization, that is an induction based of one or few examples, can be positive and favor the
fitness of the reasoner, even if it is a violence against the correct reliable way of seeing inductions,
that it is unacceptable in the light of good rational criteria for a good induction [21].
6 This phenomenon not only concerns natural language, but also emotions, actions, and all kinds
of other non propositional cognitive activities, such as the ones based on models(icons, diagrams,
visualizations, simulations, etc.).
Moral Bubble Effect 81
In this perspective we can say that moral bubbles are a perfect psychological device
that allow human beings to legitimize and the same time to dissimulate violence.
Moral bubbles, that typically characterize individual agents, can also easily charac-
terize entire groups of people that share common tasks. Usually structural violence
often appears to the agents as morally legitimate: parents, relatives, policemen, teach-
ers, and other agents in many cases perform physical or invisible violence on the basis
82 L. Magnani
of legal and/or moral reasons. Actually these legitimizing reasons cannot cancel the
reality of the violence committed and the fact that some subjects feel offended, but
surely can conceal the perpetrated violence to the actors. The concept of moral bub-
ble is consequently able to explain why so many types of current everyday violence
are considered as if they were something else.
Dowd [5] says that moral hazard is happening when one party is responsible for
the interests of another having an incentive to let his interests come first. A famous
case of moral hazard is the one regarding the subprime scandal and the related
financial crisis of fall 2008. Banks were selling their mortgages disregarding the
probabilities of default, so they misperceived all the risks at stake. I think that moral
hazard basically consists in a deceptive/faking dimension in which cooperation may
stop. To explain this aspect the following example can be useful. Humanitarian
military help to prevent genocides is considered good in western countries and by UN.
Kuperman [11] expressed a doubt about some aspects of the impact that humanitarian
intervention can have. He said that some minority leaders may have an incentive to
act without responsibility, and so to abandon cooperation, for instance provoking
bloody events to produce a good opportunity for the military intervention of the
international community on their side. Indeed, Kuperman notes that in this case it
happens that the intervention arrives too late, and the costs of the whole process are
too high. In this case the free-riding aspect of the moral hazard remains concealed
to those who have it: the minorities think they are cooperating when their are not,
because they perverted the cooperation, still favoring a moral embubblement.
Hannah Arendt splendidly noted that normal, banal, decent people perpetrate violent
and more or less atrocious acts, and this seems
[…] to go against one’s basic understanding of the world. Some exceptional explanation
must be required, because it seems that evil deeds should be done by evil people, and yet
many such deeds are committed by people who do not conform to the stereotypes of evil.
Yet these stereotypes are one of the major obstacles to understanding evil. This is ironic,
because the myth of pure evil was constructed to help us understand evil – but it ends up
hampering that understanding. The myth is a victim’s myth, and there is often a wide, almost
impassable gap between the viewpoints of victims and perpetrators [3, p. 379].
Moral bubbles are still at play. Arendt’s banality of evil is a another facet of the
them, acting at both the individual and collective level. Evil is banal because it is not
perceived as such.
Moral Bubble Effect 83
Seeing mobbing behavior in action immediately shows that civil and decent people
who usually follow modern morality, inspired by legal/civil rules—that of course
exclude mobbing—can imperceptibly adopt a mobbing behavior, simply and spon-
taneously. Mobbing is related to a kind of protomoral mechanism that Girard [7, 8]
wonderfully described as the “scapegoat mechanism”. In this case moral bubbles
render the mobber able to avoid self-contradiction and to stay calm and satisfied
with having adopted the supposed morality—momentarily considered as such—of
gossiping, obviously felt as exempt of violent possible effects.
3.7 Narcissism
3.8 Sacrifices
It is well-known that in sacrificial ritual events in which human beings and animals
were in various cases killed or harmed, violence was disguised thanks to the active
role played by the accompanying moral bubble in which the related collective was
entrapped. Indeed sacrifices are usually endowed with primary and fundamental
moral or moral/religious strong worths.
When explaining above the mobbing behavior we said that it happens thanks to a
kind of moral disengagement: civil and decent people who follow modern morality,
inspired by legal/civil rules, become violent perpetrators of possible violent mobbing.
These “disengagers” are still embedded in a moral bubble in the following sense: as
I already observed in my book [15, chapter five] the disengagement of morality—
that is at the same time the reengagement of another moral system of rules—is
performed to the aim of making possible actions that would result violent in the
light of the previous morality. In this perspective the disengagement of morality is
84 L. Magnani
Other cases could be illustrated, but here there is no room to further detail them. I
can list the most important ones.
Kant himself in [9, pp. 58–59] adumbrates the role of what I called moral bubbles
(and so I contend he corroborates my perspective) when he sees human moral situa-
tions in which the moral agent falls into a perfidious and self-deceitful reengagement
in a new decisional framework where evil is simply supposed to be good, and so
morally justified:
This is how so many human beings (conscientious in their own estimation) derive their peace
of mind when, in course of actions in which the law was not consulted or at least did not
count the most, they just luckily slipped by the evil consequences; and [how they derive]
even the fancy that they deserve not to feel guilty of such transgressions as they see other
burdened with, without however inquiring whether the credit goes perhaps to good luck, or
whether, on the attitude of mind they could well discover within themselves, they would
not have practiced similar vices themselves, had they not been kept away from them by
impotence, temperament, upbringing, and tempting circumstances of time and place (things
which, one and all, cannot be imputed to us) (p. 60).
Other cases of moral bubbles in action can be seen in the extreme behavior of
the violent “pure-evil” perpetrators, in confabulation, in the so-called fatal love, in
violent ideologies, in the so-called fascist state of the mind, in the psychological idea
of hidden trauma, in religious cognition, in the politically correct environments, and
in the moral strategies of silence, I have sketched in my book [15] and that I will
deepen later on in a further article.
4 Conclusion
of violence typical of moral bubbles are at stake. I tried to show that the concept
of moral bubble provides an integrated and unified perspective able to shed novel
light on various situations in which morality and violence are intertwined: from the
invisibility of structural violence to the problem of the so-called moral viscosity,
from the smartness of moral hazard to the difficulties of the identification between
violence and pure evil, to the issues regarding banality of evil, mobbing, narcissism,
sacrifices, and moral disengagement.
Acknowledgements Parts of this article were originally published in chapter three of L. Magnani,
Understanding Violence The Intertwining of Morality, Religion and Violence: A Philosophical
Stance, Springer, Heidelberg, 2011.
References
1. Arfini, S.: Ignorant Cognition: A Philosophical Investigation of the Cognitive Features of Not-
Knowing. Studies in Applied Philosophy, Epistemology and Rational Ethics, vol. 46. Springer,
Switzerland (2019)
2. Arfini, S., Magnani, L.: Cognitive autoimmunity: knowledge, ignorance and self-deception.
Logic J. IGPL 24(4), 612–627 (2016)
3. Baumeister, R.: Evil. Inside Human Violence and Cruelty. Freeman/Holt Paperbacks, New
York (1997)
4. Burton, R.A.: On Being Certain. St. Martin’s Press, New York (2008)
5. Dowd, K.: Moral hazard and the financial crisis. Cato J. 29(1), 141–166 (2009)
6. Gabbay, D.M., Woods, J.: The Reach of Abduction. A Practical Logic of Cognitive Systems,
vol. 2. North-Holland, Amsterdam (2005)
7. Girard, R.: Violence and the Sacred [1972]. Johns Hopkins University Press, Baltimore, MD
(1977)
8. Girard, R.: The Scapegoat [1982]. Johns Hopkins University Press, Baltimore, MD (1986)
9. Kant, I.: Religion within the Boundaries of Mere Reason [1793]. Cambridge University Press,
Cambridge (1998) (translated and edited by Wood, A., di Giovanni, G.: Introduction by R. M.
Adams)
10. Kruger, J., Dunning, D.: Unskilled and unaware of it: how difficulties in recognizing one’s own
incompetence lead to inflated self-assessments. J. Pers. Soc. Psychol. 77(6), 1121–1134 (1999)
11. Kuperman, A.J.: The moral hazard of humanitarian intervention: lessons from the Balkans. Int.
Stud. Q. 52(1), 49–80 (2009)
12. Lahti, D.C., Weinstein, B.S.: The better angels of our nature: group stability and the evolution
of moral tension. Evol. Hum. Behav. 2, 47–63 (2005)
13. Magnani, L.: Abductive Cognition. The Eco-cognitive Dimensions of Hypothetical Reasoning.
Springer, Heidelberg/Berlin (2009)
14. Magnani, L.: Understanding Violence. The Intertwining of Morality, Religion, and Violence:
A Philosophical Stance. Springer, Heidelberg/Berlin (2011)
15. Magnani, L.: The Abductive Structure of Scientific Creativity. An Essay on the Ecology of
Cognition. Springer, Cham, Switzerland (2017)
16. Magnani, L., Arfini, S., Bertolotti, T.: Int. J. Adv. Intell. Paradigms 8, 327–342 (2016)
17. Muñoz, S.A.: The nature of epistemic feelings. Philos. Psychol. 27(2), 193–211 (2013)
18. Muñoz, S.A.: Metacognitive feelings, self-ascriptions and mental actions. Philos. Inq. 2(1)
(2014)
19. Racamier, P.C.: L’inceste et l’incestuel. Éditions du Collège de psychanalyse groupale et
familiale, Paris (1995)
86 L. Magnani
20. Thom, R.: Esquisse d’une sémiophysique. InterEditions, Paris (1988) (trans: Meyer, V.: Semio
Physics: A Sketch. Addison Wesley, Redwood City, CA (1990)).
21. Woods, J.: The Death of Argument. Kluwer Academic Publishers, Dordrecht (2004)
22. Woods, J.: Errors of Reasoning. Naturalizing the Logic of Inference. College Publications,
London (2013)
The Computational Challenge
of Amartya Sen’s Social Choice Theory
in Formal Philosophy
G. Basti (B)
Faculty of Philosophy, Lateran University, 00184 Rome, Italy
e-mail: basti@pul.va
A. Capolupo · G. Vitiello
Department of Physics, University of Salerno, 84084 Fisciano, Italy
e-mail: capolupo@sa.infn.it
G. Vitiello
e-mail: vitiello@sa.infn.it
significant for making computationally effective Sen’s SCF theory, because both
based on a dynamic and not statistical weighing of the variables for interacting sys-
tems, respectively in the physical and in the social realms.
1 Introduction
The different theories depend then on the different sets of data on which the choices
are performed and, overall, on the different aggregation criteria of individual choices
The Computational Challenge of Amartya Sen’s Social Choice … 89
into one only social choice (see Sect. 3.2). Even though the precursor of the theory
of social choice is universally identified with Marie Jean Antoine Nicolas de Caritat,
the Marquis de Condorcet (1743–1794), and with his famous paradox of voting (see
below and especially Footnote 4), the social choice theory started in the last century
with the pioneering contributions of Abram Bergson and of the Nobel Laureate in
Economics, Paul Samuelson leading to the so-called “Bergson-Samuelson SWF” [3,
4], but it received a substantial improvement by the contribution of another Nobel
Laureate in Economics, Kenneth Arrow, who, for the first time, gave theory an
axiomatic form, demonstrating some fundamental theorems [5], and then finally
with a further Nobel Laureate, Amartya Sen, who gave the formalism of social
choice theory an essential contribution [6, 7] (see Sect. 3.2).
Arrows SWF is intended as a function ranking social states as less, more, or
indifferently desirable, for every pair of them, with respect to individual welfare
preferences. In the standard formalism inaugurated by Arrow, we have to do with
linear orders because based on the standard notion of transitivity that is a linear
relation.1 Formally, following Endriss ([2], p. 338):
Definition 1 (Social Welfare Function) Let L(X ) denote the set of all linear orders
on X . A profile R = (R1 , . . . , Rn ) ∈ L(X ) N is a vector of linear orders (i.e., pref-
R
erences), where Ri is the linear order supplied by individual i. We write {N(x>y) } to
denote the set of individuals that rank alternative x above alternative y under profile
R. (…) A SWF is a function F : L(X ) N → L(X ) that maps a profile of preference
orders to a single (collective) preference order.
This strict relationship so instituted by Arrow between the social choice theory and
the order theory in logic—as Endriss and Sen noticed—historically depends on the
fact that Arrow followed in 1940, as an undergraduate student, a course with Alfred
Tarski. This justifies the usage of the lexicographic (see Footnote 1) symbolism in
social choice theory, like in order theory of logic and mathematics. This formalism
is characterized by the usage of alphabetical ordering or of numerical ordering in the
symbolism, where, however, numbers are properly ordinal and not cardinal numbers.
From that, the usage of the symbol ≻ and derivatives (“higher than”) instead of >
(“greater than) and derivatives, since we have to deal here with “topological” and
not “geometrical” algebraic relations, because in principle independent of a given,
shared metrics, on which only a quantitative ordering relation can be defined.
1 Indeed, using the standard ideographic symbolism, e.g., the alphabetic letters, x, y, z, for repre-
senting relations R, the ordering relation among objects using the transitive relation is represented
as follows: ((x Ry ∧ y Rz) → x Rz). As we see, in such a way the transitive relation is linear, and
then in set-theoretic ordering it supposes a total ordering among sets. That is, for all sets the or-
dering relation ≥ holds—because admitting gaps (in our case, between x and z) among sets that
can be indefinitely long. Another possible transitive relation is the so-called “Euclidean relation”:
((x Ry ∧ x Rz) → y Rz ∨ z Ry) that has a tree structure in which no gap among sets is admitted,
and in which therefore set total ordering is not necessary, but only partial orderings are sufficient,
because in this case—and this is the essential point!—transitivity is not linear. As we see, this dis-
tinction between total and partial orderings in social choice rankings is essential for understanding
the novelty of Sen’s approach to social choice theory as to the classical ones based on a liberal
approach in economy.
90 G. Basti et al.
Given this definition of SWF as the starting point of the social choice theory, and
emphasized its relationship with formal logic and then with formal philosophy, we
can outline the scheme of our contribution.
In the Second Section of our contribution we illustrate the passage from the SWFs
to the Social Choice Functions (SCFs), in the framework of social choice theory,
in relationship with Sen’s “SCF theory”, because the passage from SWFs to SCFs
depends largely on Sen’s original contribution.
Indeed, another theoretical background to take into account in our examination of
the early social choice theories as SWF theories, is their development in the frame-
work of an approach to the study of social and economic systems at equilibrium (or in
close-to-equilibrium) conditions, as the standard formalism of statistical mechanics
in physics requires. The so-called classical and neo-classical economic theories, in
which early SWF theories were developed, are all inspired, indeed, by Samuelson’s
general approach to the mathematical economics in his seminal handbook [4]. It uses
indeed as its reference physical paradigm for modeling economic theories Gibbs’
statistical mechanics, and gas thermodynamics, to which the first two chapters of
Samuelson’s handbook are significantly dedicated. This is because this mathematical
theory is naturally consistent with the liberal individualistic vision of economy and
society. In both approaches indeed, statistical mechanics and statistical economy in
its liberal interpretation, the relations among individuals, either physical particles
or economic actors, are considered as ultimately irrelevant. And significantly, Sen’s
criticism to Samuelson’s approach starts just from this point.
Unfortunately, indeed, a fundamental unexpected and undesired consequence of
Arrow’s SWF theory is the famous “Arrow’s impossibility theorem” [5] demonstrat-
ing formally—following the efficacious synthesis of Amartya Sen—the “impossibil-
ity of the Paretian liberal” [8] for a formally consistent social choice theory. In fact,
one of the more troubling corollaries of Arrows impossibility theorem is the incon-
sistency for democratic systems of social choices based on the majority decisions
(see Sect. 2.3) on a liberal basis. That is, any liberal approach to SWF theory like
Arrow’s one takes as an axiom the so-called Pareto’s efficiency principle, for which
all the individual preference rankings have the very same possibility of determining
the social one: an abstract principle that is evidently unrealistic in photographing the
effective situation in social sciences. On the other hand, seen from the standpoint
of social choice theory, the voting democratic principle is nothing but a way for
aggregating individual preferences into one only social ranking using as aggrega-
tion criterion the majority principle. The fierce debate triggered by Arrows results
in the next fifty years contributed definitely to the development of the formal and
mathematical theory of social choices, as a fundamental mathematical discipline for
social, political, and economical sciences. Not casually, an impressive testimony of
this debate is given by the bibliography of the second enlarged edition, published in
The Computational Challenge of Amartya Sen’s Social Choice … 91
2017 [7], of Amartya Sen’s fundamental book Collective Choice and Social Welfare,
firstly published in 1970 [6].
Now, as we discuss briefly below, one of the most significant contributions of
Sen’s theory, determining its immediate diffusion among scholars and the wider
public, was precisely the formal demonstration that the only way for avoiding Arrows
impossibility results is to enrich Arrows theory with two new features.
1. Before all, by enriching SWF’s ranking (ordinal) measures with (cardinal) mea-
surements and comparisons of capabilities, resources, and utilities among in-
dividuals, so to transform SWFs in as many and more realistic social welfare
functionals.
2. Secondly (and definitely), by introducing into the model the interpersonal com-
parison of goods and utilities—and, more generally, the information exchanges
among persons constituting homogeneous groups—, on the contrary considered
as irrelevant in the classical (liberal) economic theory, as well as in standard SWF
theories, like Arrows one.
In the Third Section of this contribution we discuss therefore the main characters of
Sen’s “social choice function (SCF) theory”. Indeed, the two just remembered im-
provements as to Arrows SWF theory allowed Sen to introduce into the mathematical
modeling of his SCF theory comparative principles of social and economic equity
or fairness, using as aggregation principle the ethical maximin principle that gives
priority to the interests of worst-off persons used before by one of Sen’s teacher,
John Rawls, in his celebrated political theory of “justice as fairness” [9]. The main
difference, as we see, is that, for Rawls’ neo-Kantian normativism the maximin prin-
ciple supposes again—if applied to social choice theory—an untenable in our plural
society, because absolute and then abstract, “total ordering of the distributive justice
grading. On the contrary, this limit does not hold in Sen’s SCF theory, where only
partial orders are allowed, but where it is possible to merge different local gradings
of justice into one shared ranking. In such a way, the maximin becomes a powerful
aggregation criterion of comparative distributive justice, not supposing any unreal-
istic reference to an absolute criterion, for unveiling and solving concrete injustice
situations. This possibility transforms Sen’s SCF Theory into a comparative theory
of social choices, which considers as intrinsic to the models also the ethical and the
interpersonal constraints, because of Sen’s wider and systematic usage of the max-
imin principle as a local equity principle we illustrate below. On this regard, from the
standpoint of the formal philosophy of ethical choices, the fundamental contribution
of Sen was the formal demonstration that an effective mathematical modeling of
ethical constraints in economy cannot be based on abstract and not-computable op-
timal choices defined on a unique, because supposed as complete (=total) grading of
distributive justice for social/economical states in a society, but on concrete criteria
of maximal choices relative to the different contexts, that is, based on interpersonal
equity criteria. This mathematical modeling of SCFs can constitutes also the basis
for more suitable measures of social well-being and welfare than GDP, recently de-
veloped in social, political and economic sciences with the passionate and exciting
public discussions that followed, to which Sen gave a fundamental contribution [10].
92 G. Basti et al.
Finally, in the Fourth Section of this contribution we want to discuss the computa-
tional challenges of Sen’s SCF theory. Indeed, as Sen himself emphasizes in many
places, precisely the difficulty of applying his theory using the classical statistical
tools of the economic analysis is the main “technical” obstacle—apart from the ideo-
logical ones—against its systematic application. The problem is well synthesized by
Sen saying that an approach like his SCF theory, all based on the interactions among
personal and collective social actors, requires from the mathematical standpoint a
dynamic weighing of the variables that no statistical weighing function can grant in
principle.
Moreover, also from the computer science standpoint, we are faced in AI machine
learning with the very same problem. As we see, indeed, it is absolutely unfitting with
the required dynamic weighing of the variables the statistical machine learning model
in AI systems, based on the so-called “deep-learning” approach of artificial neural
networks (ANN) in AI (see Sect. 4.1). This approach, indeed, maintains the classical
distinction of the statistical analysis, between the determination of the statistical
weights of the variables, and their further application on the whole data set. In the
ANN modeling, this distinction becomes between:
1. The necessarily automatized—because of the presence of inner higher order cor-
relations among huge amounts of data that makes impossible the determination of
the weights by hand—(deep) learning phase for fixing the weights of the variables
on a representative sample of the data set, followed by
2. The testing phase, consisting in the application of this statistical weighing to
another different representative sample of the data set, and then, if this testing is
successful, to the whole data set.
To stress the relevance of this point, let us consider the same problem of the infor-
mation exchange among social/economical actors from another but complementary
standpoint as to Sen’s political theory. In fact, the real-time information exchange
among communicating agents that internet today allows, determines the fast aggre-
gation/dissolution of interest groups in a world-wide environment. For an effective
representation of this phenomenon evident to everybody, think, for instance, at our
worldwide currency and stock-exchange markets, and at the streams of data, with
always changing inner correlations, which internet produces in real-time,2 before
on the computer screens, and then on the consequent choices of the financial opera-
tors, distributed over the five continents. As we know, this problem of “infinite data
streams” in Theoretical Computer Science (TCS) belongs to the wider class of the so-
called “big-data issue”, determining the growing interest to innovative computational
solutions, the controversial quantum computer architectures included.
Indeed, “infinity” is here related with the continuous, sudden aggregation/
disaggregation of coherent patterns of behavior of the financial agents (humans and
computers), in markets distributed all over the world, dramatically represented by the
2 E.g., think at Bloomberg data streaming, updating continuously the values of shares and of curren-
cies on the worldwide financial market, active all over the day because of the different time zones,
and that are changing in average every 10 s.
The Computational Challenge of Amartya Sen’s Social Choice … 93
negates the universality of human rights, and so offends the weaker persons and
peoples in our society, without, on the other hand, supposing an absolute ethical
system of reference that would be again another form of oppression.
Indeed, by substituting the abstract universalism/normativism of a theory of jus-
tice, such as, for instance the Kantian one, with a concrete and computationally
effective theory of “comparative justice as equity”, because based on criteria of
equity among persons and groups sharing different value and moral systems, it is
possible to make effectively and universally comparable the local solutions of global
problems for different contexts in our open society.
This is a tremendous task [14] for which the aid of suitable AI systems for sup-
porting the social choices of policy and decision makers in economics and society
is unavoidable, according to a notion of collective wisdom, including necessarily
humans and machines [15, 16]. All this emphasizes the relevance of proposing com-
putationally effective models in the field of social choice theory. And these reflections
will constitute the Conclusions of our contribution.
Generally, in economic literature, there are two main types of Social Welfare Func-
tions (SWFs), as far as they are defined respectively, either on
1. Some support (domain-codomain) of real valued economical magnitudes (defined
on cardinal numbers) for one only social group, or
2. Orders, i.e., domains-codomains of rankings (ordered sets) of preferences, and
of rankings (ordered sets) of social states, both defined obviously on ordinal
numbers.
Now, when we are speaking about orders, effectively we are considering a SWF
like a particular type of Collective Choice Function (CCF), relating ordered sets
of individual preferences/utilities, with ordered sets of social/economical states for
each individual. This means that, mathematically, we are moving from real valued
(cardinal) functions, to set-theoretic logic functions—effectively we are moving to
set-theoretic semantics (“order theory”) applied to social entities. The initial rep-
resentatives of these two types of SWF theories generally quoted in literature are,
respectively, “Bergson-Samuelson SWF” and “Arrow SWF”.
The Computational Challenge of Amartya Sen’s Social Choice … 95
Abram Bergson first introduced in economics the SWF notion as real-valued dif-
ferentiable functions, aimed at formally representing “the conditions of maximum
economic welfare” for the society as a whole. Bergson’s SWF then includes as func-
tion arguments several quantities of different commodities produced and consumed,
and of resources, labor included [3].
The fundamental contribution of the 1970 Nobel Prize in Economics Paul
Samuelson—founder of the prestigious “MIT School of Economics”, counting
among its members an impressive list of Nobel Prizes—is synthesized in what is
commonly called the “Bergson-Samuelson SWF”. It aims at representing (in the
maximization calculus) all real-valued economic measures of any belief system—
“that of a benevolent despot, or a complete egotist, or all men of good will, a mis-
anthrope, the state, race, or group mind, God, etc.”—required to rank consistently
different feasible social configurations as “better than”, “worse than”, or “indifferent
to” each other in an ethical sense (see [4], p. 221).
What is essential for our aims is that Samuelson’s modelling of the equilibrium
stability for economic systems explicitly depends on Willard Gibbs statistical me-
chanics interpretation of thermodynamic systems, as Samuelson explained in the first
two chapters of his masterpiece Foundations of Economic Analysis, of which eighth
chapter is dedicated to welfare economics and then to his SWF interpretation in such
a theoretical framework (see [4], pp. 203–256).
The second type of SWF is related with the work of Kenneth J. Arrow, awarded with
the Nobel Prize in economics on 1972. Since the first version of his theorem (1948),
indeed, he transformed effectively Bergson’s SWF into a collective choice function
(CCF), as the same title of his fundamental book Social Choice and Individual Value
exemplifies [5]. That is, whereas the Bergson-Samuelson SWF rules the mapping
from any set of individual orderings of preferences/utilities into one only set of
ordering of social states, Arrow SWF rules the mapping from any set of individual
orderings of preferences/utilities into a set of social states, among many alternative
ones (see [5], pp. 22–25). As Arrow himself emphasizes the interpretation of SWF as
a CCF is effectively a restriction over SWF, because it requires that for any individual
ordering, for some sufficiently wide but finite range of them, the SWF “give rise to
a true social ordering” among finitely many ones.
Arrow’s CCF effectively poses other conditions to SWF giving rise to the famous
Arrow Impossibility Theorem. Following the later version of Arrow’s theorem (see
[5], pp. 46–61) these conditions can be synthesized as follows.
96 G. Basti et al.
3 There exists evidently, and many Authors noted it, a relationship between Arrow’s impossibility
theorem and the famous Condorcet’s paradox on voting stated in his Essai sur l’Application de
l’Analyse à la Probabilité des Decisions Rendues à la Pluralité des Voix published in 1785 [17],
and considered unanimously as the official birth-date of the modern, mathematical theory of social
choices. However, Sen demonstrated that Arrow’s impossibility theorem cannot be reduced to
Condorcet’s paradox, before all because it would be hard to argue that majority rule would really be
a plausible way of aggregating preferences in welfare economics (see [7], p. 275). See also [7], pp.
395–419, and overall the concluding chapter of Sen’s book ([7], pp. 453–472), in which he offers a
deep analysis of what a modern comprehensive, because rational theory of social choice requires in
our “Communication Age”, and that is far beyond any oversimplified and oversimplifying “majority
criterion”.
The Computational Challenge of Amartya Sen’s Social Choice … 97
Effectively, two are the main ways for avoiding the outcome of Arrow’s theorem:
1. Working on infinite sets of preferences—i.e., on finite sets of ever-changing pref-
erences and aggregation rules—, and not on finite (static) ones like in Arrow’s
SWFs. In such a case, however, as far as we suppose a total ordering relation
of sets, the aggregation rules seem to be of limited interest because in standard
set-theoretic logic they are based on ultrafilters,—i.e., maximal filters on given
“partially ordered sets” or posets (see Rel. 9 below), supposing the total order-
ing (see Rel. 8 below) of all partial orders. Namely, this means a supposition of
completeness in the definition of the social rankings that is simply illusory to
suppose in complex and global societies like ours. “Ultrafilters”, in other terms,
are non-constructive mathematical objects because they suppose the (infinite)
set total ordering—i.e., their definition requires a non-computable higher order
logic—, so to require practically as many “invisible dictators” [18]. And, in fact,
such aggregation rules violate Turing computability [19], so to result practically
useless in mathematical social choice theories, and in social computing applied
to economics and to social sciences.
2. Working exclusively on set partial orders in social choice theory, instead of the
strict orders of Arrow’s SWF (see Sect. 2.4 the list of the main set ordering
relations), so to define another type of rational CCFs, the Social Decision Function
(SDF). For SDFs Arrow’s impossibility theorem, indeed, does not generally hold,
even though it holds for SWFs as special cases of SDFs. This approach was firstly
pro-posed by another Nobel Prize in Economics (1998), Amartya Kumar Sen,
who gave in this way an essential contribution of clarification to our discussion
[6, 7].
We can anticipate, however, that also in Sen’s approach the challenge of infinity and
hence of effective computability on infinite sets, e.g., data streams, represents itself
as an unsolvable computational problem within the Turing Universality paradigm
in computations. This holds also in the case of the “probabilistic” Turing Machine”
architecture, artificial neural networks (ANN) included, as far as based on a purely
statistical weighing of the variables, through the machine learning principle—i.e.,
the so-called “deep learning” of the actual AI systems—of the stochastic gradient de-
scent, as we discuss briefly in the Third Section of this contribution. These limitations,
indeed, hold till we are working in standard set theory—e.g., the Zermelo-Fraenkel
(ZF) set theory—and not in non-standard ones, like the theory of Non-Wellfounded
(NWF) sets, requiring the “Algebraic Universality” in computations, and not the
“Turing Universality” that is a subset of the former one, as we hint below. Unfortu-
nately, a satisfying discussion about this fundamental point would require another
specific study that we cannot develop in this contribution.
98 G. Basti et al.
(see [6], p. 10). Then, Sen notices, in the case of Arrow’s SWF, the ordering must
effectively satisfy 1, 2, and 3, but it is, per se, irrespective of 4 and then of 5 (see
[6], p. 8). In such a way, we can avoid, under some conditions explicated by Sen
in the rest of the book that we only summarize here, the uniqueness of the pivotal
individual for all social orderings, as depending on some rational CCF on social states
he defined as social decision function (SDF). This opens the possibility of satisfying
simultaneously the conditions P* and D* of Arrow’s theorem as characterizing any
CCF, and then a SDF too, even though at the price of an organization of social states
much more fluid, than in Arrow’s vision.
In other terms, the core of Sen’s approach consists in demonstrating formally
that we can obtain suitable CCFs such as SDFs, satisfying all Arrow’s conditions
without falling into his “impossibility result”, by supposing quasi-orderings and
overall partial orderings instead of strict orderings like in Arrow’s approach. We
refer to ([7], pp. 313–337), for a complete discussion of Sen’s formal treatment of
the CCFs and SDFs in social choice theory.
Anyway, to conclude this subsection, let us recall, another interesting consequence
deriving from the finiteness of SDFs—and of the Social Choice Functions (SCFs)
we discuss below.
That is, the possibility of representing in it also judgements about values, shared
by a homogeneous social group because characterized by strong interpersonal ex-
changes of information. These topics were the main object of a more recent essay
of Sen, with the significant title: The Informational Basis of Social Choice [20], and
they are the leit-motiv of the new part added to the Second Expanded Edition of [6],
even though largely discussed also in ([6], pp. 89–117), where he introduced the
transition from a SDF to a SCF. We return in the next subsection on this fundamental
transition.
However, when we study social and economic systems, taking into account also
the motivational social forces acting within them, overall if spread in real time by
social media, the challenge of infinity represents itself because making untenable that
either a SDF or a SWF can range over a fixed finite set of social state alternatives,
chosen by individual preference orderings.
Nevertheless, the situation changes radically when we pass from individual pref-
erence orderings (rankings), to social choice functionals, because expressed by ho-
mogeneous groups of persons, that is, by groups of individuals characterized by a
strong mutual information exchange making them an unique social subject: this is the
core of the passage from SDFs, to SCFs. Using a thermodynamic metaphor, this is
the core of the passage from a “liberalist” vision of society seen as a “gas” or “vapor”
of individuals, to a “pluralist” vision of society characterized by the “condensation”
of individuals into homogeneous but reciprocally different groups—like as many
different “drops”, for continuing in the thermodynamic metaphor. In such a case,
the “infinity challenge” represents itself in a different way, because of the boundary
condition of the social media influence, often on a world-wide dimension—think
at internet, before all. The influence of social media indeed, in specific contexts,
acts like sudden changes of “temperature” in the human environment, determining
immediate rarefaction/condensation processes of old/new groups inside the human
100 G. Basti et al.
In this section we develop our analysis in several steps. After a general definition of
what is a social choice function (SCF) in social choice theory, we examine, before
all, Sen’s social choice function definition, and the consequent “SCF possibility
theorem”.
Afterward, we discuss the relevance of the information basis in SCFs. Then, we
examine John Rawls’ theory of “justice as fairness” in the context of social choice
theory because this prepares Sen’s comparative theory of justice as equity, on which
his SCF theory is based, and that constitutes the core of its originality and complexity.
The examination of the anthropological background of this theory constitutes
particularly the object of Sect. 3.4 concluding this section. Indeed, Sen’s SCFs theory
strictly depends on his original notions of the so-called personal capabilities and
functionings, as the key-concepts of his comparative theory of “justice as equity”.
It is evident that a SCF escapes from Arrow’s impossibility results simply because
Arrow conditions are essentially binary relational demands (i.e., individuals vs. one
social state ranking). In fact, in the present case, what is required is not a SWF (or
SDF) (see Sect. 2.4), but a SCF:
g (S, {Ri }) such that g (S, {Ri } ⊆ S for all non-empty subsets S ⊆ X (2)
This is essentially equivalent to making the value of the function f (Ri ) a finitely complete
choice function C(•) for each society, and not – as with SWFs or SDFs – a social preference
relation R [7], p. 315.
In this way, for such a FCCR: f (•), the SCF theory requires a re-definition of Arrow’s
conditions U, P, D and relative axioms (see Sect. 2.3), in terms of restrictions of n-
ary choices over pairs only. That is, Û , P̂, D̂, plus the extension of non-dictatorship
condition
! ′" D to a stronger #!
condition of “anonymity” Â. I.e., if Ri is a permutation of
′ "$
Ri , then f ({Ri }) = f Ri .
Unfortunately, we cannot discuss here this significant point, and we refer to [7], p.
316 for further particulars. Now, if we combine these conditions to the other Arrow’s
condition I of “independence of irrelevant alternatives” that was defined in functional
terms, i.e., as an I A by Arrow himself, we can state the following theorem [7], p.
316:
Theorem 1 (Choice-functional Positive Possibility Theorem) For H -ariet y ≥ 2,
there is a FCCR satisfying conditions Û , I A , P̂, D̂ and Â.
In the following discussion [7], pp. 316–336, Sen emphasizes the relevance of
having an ariety (number of function arguments) H -ariet y ≥ 2, where H is the set
of individuals, in a social choice functional interpretation of SWFs and SDFs. Indeed,
also in the case of restrictions of choices over pairs only, a FCCR must satisfy anyway
the condition of transitive closure x B ∗ y of a sequence of at least a triple of binary
relations xBy, having necessarily a maximal choice set C(S max ) including the factor
BA of B, defined as “xBy and not yBx”, which is an asymmetry condition that closes
the sequence, avoiding any cyclicity. We return in Sect. 3.4 on the centrality of this
maximality condition for emphasizing the computability and then the effectiveness
of Sen’s SCF theory also if we inserted in it ethical personalistic components.
Anyway, this can be seen straightforwardly also in the more elementary case of
Condorcet’s paradox of voting, interpreted as a choice function. In fact, the transitive
102 G. Basti et al.
4 Indeed, the paradox derives from the fact that voter 1, 2, and 3 are ranking the three alterna-
tives in the orders: x, y, z; y, z, x; and z, x, y, respectively. In this way, there is a preference
cycle, and each alternative is defeated in a majority vote by another alternative. Now, if we do
not consider preference maximization, but maximality in terms of the transitive closures of pref-
erence relations, we have the following choices: C ({x, y}) = {x}; C({y, z}) = {y}; C({z, x}) =
{z} ; and C {(x, y, z)} = {x, y, z}. Of course, this choice function is not binary, but satisfies all the
Arrow axioms in their choice-functional version, of which at Theorem 1 (see [7], p. 316).
The Computational Challenge of Amartya Sen’s Social Choice … 103
attaches relevance to the economical utilities of individuals, and it does not consider
information about fulfilment or violation of rights or of liberties, levels of incomes
or affluence that people enjoy—if not for the indirect effects that such variables have
on individual utilities. On the contrary, a libertarian theory of justice stresses over
individual liberties and rights, and it gives no direct relevance to the levels of utilities
or of incomes, and of their interpersonal comparison.
To sum up, Sen defines the informational bases of different theories of justice for
inserting them in the functional structure of a SCF in terms of (see [7], pp. 339–340):
1. The basal space of the theory, i.e., the class of variables on which the justice
assessment of that theory depends (e.g., in utilitarian theories, the combinations
of the individual utilities);
2. The aggregation system of those variables, for allowing a discriminating use of
these variables. E.g., in utilitarian theory, the individual utilities are simply added
together to arrive at a sum-total, serving as basis for the assessment of the so-
cial state, without paying attention, for instance, to some measures of dispersion
or inequality. In this context, the aggregation systems of theories of justice that
include as their essential components “distributive criteria” of commodities, utili-
ties, and opportunities among individuals become particularly relevant. These are
the so-called theories of justice as fairness or equity like Rawls’ and Sen’s ones.
Effectively, in all his work and not only in his Collective Choice book in both its
editions (see [6], Chaps. 9 and 9*, and [7], pp. 187–219; 347–364), Sen pays a
particular attention to the Theory of Justice as Fairness of one of his former teachers,
Rawls [9, 21–24].
This theory influenced deeply not only Sen’s theory, but the political philosophy
of XX cent. overall in US, in the sense of a “moderate liberalism”, if not in the sense
of a “social-democratic liberalism”, even though such labels are not able to do justice
of the relevance, the richness and the originality of this theory.
Indeed, Rawls theory of “justice as fairness”, because it introduces a fundamental
criterion of “equity” or “fairness”—that is, of interpersonal comparison of welfare—
as constitutive of the idea of justice, establishes also a substantial point of distinction
as to the Kantian formalist theory of justice. Even though Rawls wants explicitly to
offer by his theory an updated version of Kant’s ethics in the direction of a theory
of distributive justice, without losing, however, the unconditional character of the
moral principles characterizing Kant’s theory.
Not casually, indeed, in his systematic treatment of Rawls’ theory in both editions
of his book on Collective Choice (see, for instance, [7], pp. 187–209), Sen starts
precisely from the quotation of the fundamental statement of I. Kant’s Grundlegung
zur Metaphysik der Sitten (1785) [25], synthesizing his formal principle of “duty
for duty” as the unconditional basis of the “moral oughtness”, as distinct from the
104 G. Basti et al.
unconditional “logical necessity”. Namely, this principle reads: “act always on such
a maxim as thou canst at the same time will to be a universal law”.5 As we know, in
this way Kant wanted to satisfy the so-called “Hume’s law” distinguishing between
logical necessity and ethical oughtness—sometimes denoted also as the reciprocal
irreducibility facts/values—in philosophical (modal) logic. At the same time, how-
ever, he wanted to preserve the universality of moral laws by granting them the same
unconditional or absolute character of the logical laws. That is, by granting to moral
laws the same tautological character of logical laws by the principle of the “duty for
duty”, without confusing them. Historically, in this way, Kant wanted to criticize the
Aristotelian foundation of justice on equity made proper by the Scholastic Tradition,
which seems to imply a factual comparison of what is effectively right for different
individuals in similar circumstances.
Not casually the modern re-proposals of the theory of equity or fairness before
Rawls, as Sen recalls, tried systematically to reconcile the notion of equity with the
absolute universality of the Kantian moral law, by making irrelevant the personal
differences. On this regard, Sen quotes Henry Sigdwick’s approach that is paradig-
matic in this sense because perfectly compatible with Kant’s principle (see, [7], pp.
187–188).
…Whatever action any of us judges to be right for himself, he implicitly judges to be right
for all similar persons in similar circumstances. Or, as we may otherwise put it, if a kind of
conduct that is right (or wrong) for me is not right (or wrong) for someone else, it must be
on the ground of some difference between the two cases, other than the fact that I and he
are different persons. A corresponding position may be stated with equal truth in respect of
what ought to be done to – and not by – different individuals ([27], p. 374).
In the formalization of Sigdwick’s theory of equity firstly proposed by Hare [28, 29]
the universality of moral judgements consists ultimately in inserting their invariance
for permutation of persons within the similar circumstances recalled by Sigdwick,
everything else remaining the same. Of course, Sen notices, there was a lot of crit-
icisms to Hare’s modelling of equity theory as a formal basis of a social choice
theory, because it requires that a moral judgement be held from every position that
a person can occupy in a society through interpersonal permutations. The reference
to Sidgwick’s and Hare’s theory of equity is useful to Sen, however, for introducing
the originality of Rawls’ theory about justice as equity, for which no interpersonal
permutation of positions is required since for Rawls “the principles of justice are
those which would be accepted in a fair situation in the original position” ([7], p.
192).
In a paper of 1958 that anticipates his famous book of 1971, but it is often ig-
nored by the scholars, Rawls defined, indeed, the original position as “an initial
position of equal liberty (which is defined by the pattern of rights and duties, pow-
ers and liabilities, established by a practice)” ([21], p. 166), from which any theory
on the principles of justice must start. In Sen’s words, “the principles of justice
are those which would be accepted in a fair situation in the original position” ([7],
p.192). Effectively, this position is for Rawls a purely hypothetical and a-historical
These two principles are interpreted by Sen as constituting the information basis
of Rawls theory of justice. That is, respectively, as determining the basal space of
the main variables characterizing the theory (First Principle), and as determining
the “aggregation criteria” of these variables (Second Principle), mainly the maximin
principle (a).
In this framework, according to Sen ([7], p. 348ff.), the basal space of Rawls theory
is constituted essentially by the “basic liberties”, but in reference to the “primary
goods” that each individual as citizen must hold, and this is a first point in which Rawls
differentiates himself from the libertarianism. Indeed, Sen emphasizes, the range of
these liberties is narrower than in libertarianism—e.g., it does not include property
rights or rights of exchange or of bequeathal. Primary goods are, in fact, general-
purpose resources that are useful for the pursuit of objectives that the individual
may have so that, in Rawls words, they are things that citizens need as free and
equal persons, and claims to these goods are counted as appropriate claims ([30],
p. 257). “Primary goods” are then “things that every rational man is presumed to
want”, and include, without any pretension of exhaustiveness, beside the “basic
liberties”, “income and wealth”, “freedom of movement” and “choice of occupation”,
“equality of opportunities”, etc. The reference to human rationality as characterizing
all human individuals, makes evident that, in moral philosophy, Rawls primary goods
correspond to moral goals, which are proper to humans as such, and then that each
individual universally (naturally) wants, so that each person has the right to pursue
them in whichever society. Namely, they correspond to what in political and law
philosophy are defined as “human rights” that naturally must have their place in
106 G. Basti et al.
any social choice theory (see [7], pp. 420–452). On the other hand, the fundamental
aggregation principle of Rawls theory is, according to Sen ([7], p. 350), the maximin
principle. Indeed, Sen emphasizes ([7], pp. 193–194), the component (a) of the
Second Principle of justice in the “original position we quoted before and by Rawls
himself defined as the Difference Principle, because it requires that social inequalities
be arranged to make the worst-off best-off, is essentially a maximin criterion. In fact,
it requires that the minimal element in the set of individual welfare’s be maximized.
Now, despite Rawls main focus is on the theory of type of institutions to be chosen
in order to satisfy at best his two principles of justice, and not to propose a social
choice theory, nevertheless the maximin principle could be used in principle also
to order social states based on individual orderings, and so within a social choice
theory framework. For instance, for any social state, we may order the individuals
in terms of their welfare and pick on the worst-off individual. This can be compared
with the welfare of the worst-off individual in another social state, and so on, so to
have recursively, at least in principle, a complete social ordering ([7], p. 193).6
Nevertheless, Sen rightly emphasizes, the maximin principle interpreted as a sort
of choice procedure in a social choice theory is not a SWF in the sense of Arrows, since
the maximin procedure imposes that a worst-off individual i in a social state x goes
up for every alternative so to change his situation. This means that the social ordering
involving x, being based on a different individual welfare’s, becomes different. And
this is incompatible with any SWF (see [7], p. 193). Neither the maximin procedure
is a Collective Choice Rule (CCR), since a CCR specifies a social preference relation
based on the set of individual orderings of social states, each from one’s own point
of view. In other terms, a CCR is based on n orderings of m elements, whereas a
Rawlsian maximin choice mechanism is based on one ordering of mn elements (see
[7], p. 194).
At this point many problems raise in the possibility of interpreting Rawls’ maximin
principle as an effective decision rule in a social choice theory—as unfortunately
it has been done too many times by policy makers, and in recent times, by the
so-called “populist movements”. Indeed, if it is absolutized as a decision rule in
social choices, the maximin criterion becomes easily a factor of social injustice and
economical regression, given that it oversimplifies complex situations. And in fact, it
is often accompanied by misinformation campaigns on the media, which are, in fact,
a perverse way of implementing the Rawlsian principle of the “veil of ignorance”
over the real causes of the actual differences, making this veil not “voluntary” but
“imposed”.
Anyway, some of the main limitations of using the Rawlsian maximin principle
as a decision rule in a social choice theory are summarized by Sen in ([7], pp.
195–198). Ultimately, they reduce themselves to the fact that Rawls criterion of
fairness, as far as referring to a hypothetical original position that, for definition, is
irrespective of the historical, actual personal differences (covered by a voluntary veil
of ignorance), becomes a too severe condition - in some proper sense, “ascetical”,
6 Effectively,
Sen offered in [6], since its first printing in 1970, a lexicographic formal version of
the maximin principle that Rawls himself included in his book A Theory of Justice of 1971 [9].
The Computational Challenge of Amartya Sen’s Social Choice … 107
because this requires that individuals renounce voluntarily to take into account their
personal differences, for being used in a theory of decisions about effectively fair
social choices.
Indeed, we repeat, it risks becoming a false strategy of curing the symptoms of
an illness without removing the causes—or because there is the precise will of not
removing them, as it is the case of the social media manipulators behind populist
movements. Nevertheless, it remains true, as Sen emphasizes, that in several institu-
tional questions that is the proper focus of Rawls theory of justice, the appeal of the
Rawlsian maximin criterion is well demonstrated and it works. Precisely because
it cures one—if not the principal one—cause of a social disease: the existence in a
society of some unjust institutions, because not aimed, each in its proper way (e.g., as
an educational, or as an occupational, or as a political institution, etc.) at maximizing
the social states of worst-off individuals.
However, what Sen notices is that the maximin principle suggested by Rawls
can be used also independently of his theory of justice, in a social choice theory.
Particularly, in a theory of justice as equity like that implemented in Sens SCF
theory (see Sect. 3.4). It, indeed, starts from the actual personal—also subjective
and not only objective—differences among individuals, the situations of manifest
injustice or iniquity, without veiling them, before all. That is, it starts from the
relevance in a theory of justice as equity of the subjective differences (including the
ethical and cultural differences and preferences, the religious beliefs, or even the
simple personal tastes) in the interpersonal comparison of welfare states, because
determining concretely different subjective ways and grades of enjoying what are,
objectively, the same commodities, and/or of profiting by what are objectively the
same utilities.
Indeed, precisely such differences among individuals and groups in a society make
untenable the Pareto unanimity axiom, we discussed in Sect. 2.3, so to transform
Arrow’s “impossibility theorem” in Sen’s “impossibility of the Pareto liberal” [8],
because incompatible with the definition of equitable social rankings. Indeed, to sum
up the example discussed at length by Sen in ([7], pp. 207–209), if the increasing
of GDP in a given nation, e.g., in India, depended on the growth of production of
beef, the impact would be different over Muslims and Christians welfare states, with
respect to Indus one. Just as, on the contrary, it would happen for the welfare states of
Indus and Christians, as to Muslims one, if GDP increasing depended on the growth
of production of pork.
In a word, in Sen’s theory of justice as equity, the interpersonal comparison of
states and preferences for defining a common ground of equality, on which applying
the maximin criterion, is ethically based not on a voluntary “veil of ignorance” over
the differences, but on the “ethical “principle of extended sympathy in the form of
placing oneself in the position of another” ([7], p. 290]). The starting point is, in
this case, not the usual individual ordering Ri , but an extended individual ordering
R̃i, j of individuals i, j, and states x, y, one for each individual: (x, i), (y, j), where
i ̸= j. That is, the “extended ordering” is an ordering where the subjects are not
mere individuals i, only numerically distinct, but persons i.e., different individuals.
108 G. Basti et al.
Starting from this extended ordering, the principle of “extended sympathy” that
Sen borrowed from Adam Smith, as we see, is formalized by our Author within his
SCF theory in terms of two axioms of identity. The first one reads: “each individual
j in placing himself in the position of person i, takes on the tastes and preferences
of i ([7], p. 214). This ethical subjective principle can be objectively extended to the
second axiom of “complete identity”, so to give a formal criterion of grading justice
for the “interpersonal comparison and permutation” among social states of different
persons in a SCF theory.
That is, the first axiom can be extended to the identification between different
preference rankings J of the social states, x, y, for any pair of different persons i, j,
i.e., x Ji y = x J j y, where x Ji y reads: “x is more just than y according to person i”
([7], p. 210).
What is fundamental for the logic of Sen’s SCF theory, is that, while each pref-
erence ranking Ji in P. Suppes’ theory of “grading justice” [31], from which Sen
derived his own, is a strict partial ordering over the set {X } of possible social states,
if we insert the possibility of making equivalent two or more of preference rankings
originally different, each Ji becomes a “simple” partial ordering7 Oi , which is suf-
ficient for extending the Rawlsian maximin principle, interpreted as an aggregation
principle, to Sen’s SCF Theory (on this essential point, see Chap. 9* of both [6, 7]).
In a word, what in Rawls justice theory is a hypothetical condition of fairness
among individuals, hiding under a voluntary veil of ignorance the objective and
subjective actual differences among persons, in Sen theory such differences are the
starting point. In this way, fair judgements and solutions of unjust inequalities become
the final point of a process of interpersonal comparability and information exchange,
in the limit leading to an ethos of dialogue among different persons and groups, in a
multi-cultural society.
Complementarily, for any assumption of real-valued (statistical) measurability
and interpersonal comparability of personal welfare states, in a SCF theory based
onto Sen’s theory of grading justice as interpersonal equity, the axiom of complete
identity becomes operationally an equivalence between vector spaces V, V ′ of ⟨m|n⟩
alternatives. That is, of m states for n individuals for two different persons i, j, onto
which calculating the statistical expectations in a quantitative theory of real-valued
welfare functions (see [7], pp. 210–220 for the formal treatment of these notions).8
7 We recall here what we illustrated in Sect. 2.3. Namely, that a “strict” partial ordering is an order-
ing R(≤) among sets, in which only the asymmetric case ((x Ry ̸= y Rx) → x ̸= y) holds of the
antisymmetric relation that, on the contrary, holds for the partial ordering as such. That is, where
also the symmetric case, ((x Ry = y Rx) → x = y), holds.
8 The usage of the ⟨bra|ket⟩ notation for the matrix representation of Sen’s doubled vector states—
a notation made famous by the quantum formalism, even though not limited to it—is an implicit
suggestion that the more effective approach to make computationally effective the two axioms of
identity of Sen’s SCF theory is the usage of the quantum state superposition principle. In the frame-
work, however, of the doubling of the degrees of freedom formalism for calculating dynamically
the statistical expectations for entangled interacting systems, typical of the “many-body physics”
approach to QFT. That is the fundamental physics of condensed matter physics (see Sect. 4).
The Computational Challenge of Amartya Sen’s Social Choice … 109
To conclude this section, let us discuss briefly what is he proper of Sen’s theory
of justice as equity among different persons, characterized by different capabilities
and functionings, intrinsically related, objectively, to different social contexts and
related opportunities, and, subjectively, to different goals and value systems, as a
fundamental ethical component of Sen’s SCF theory.
In other terms, the core of Sen’s theory of justice as to liberal and utilitarian theories
consists in considering commodities and utilities as means for reaching those goods
or “valuable ways of being and of doing”, which a person—on the basis of her/his
system of values—considers as fundamental for living “a valuable and valued life”,
that is, a personal achieved life.
A personal achieved life can be seen as a combination of functionings (i.e., of doings and
beings), and, taken together, can be the basis for assessing that persons quality of life. The
functionings on which human flourishing depends include such elementary things as being
alive, being well-nourished and in good health, moving about freely, and so on. It can also
include more complex functionings, such as having self-respect and respect of others, and
taking part in the life of the community (including appearing in public without shame), on
which Adam Smith in particular presented an extraordinarily insightful analysis in his Wealth
of Nations ([7], p. 357).9
9 Apartfrom the reference to Smiths masterpiece, Sen quotes in note another less famous book of
A. Smith, The Theory of Moral Sentiments, of which Sen himself cured a re-edition with his own
Preface [35].
110 G. Basti et al.
in the set of combinations of functionings from which the person can choose any one com-
bination. Thus, the capability set stands for the actual freedom of choice a person has over
the alternative lives that he or she can lead ([7], p. 357).
To sum up, the “basal space” of Sen’s theory of justice is characteristically constituted
by “what people are able to be or are able to do”, rather than in terms of the means or
resources they possess. In this way, for the construction of a suitable “social choice
theory based on these personalistic principles,
individual claims are to be assessed neither just by the incomes, resources or primary goods
people respectively have, nor only with reference to the pleasures or utilities they enjoy, but
in terms of the freedom they have to choose between different ways of living they can have
reason to value ([7], pp. 357–358).
Then, in a social choice theory, of which information basal space is given in terms
of “capabilities and functionings”, the following question becomes fundamental.
Which is the aggregation principle of the different variables and of the different
interpersonal comparisons, able to take note of variations of individual conditions
and circumstances, in which even the underlying personal values and preferences
play an essential role?
As we anticipated, as far as we introduce in Rawls’ maximin aggregation principle
the two axioms of extended identity among different welfare rankings, this principle
can be extended as well to Sen’s SCF theory, as far as we can suppose a suitable
real-valued quantitative parametrization of the welfare aggregates involved (see [7],
pp. 210–220 for a formal treatment).
Indeed, by these axioms we can formalize, on the one side, the ethical principle
of the interpersonal/intercultural dialogue as “extended sympathy” among different
persons—that is, the ability of posing him/herself in the situation of the other one,
taking on also his/her personal and ethical preferences and tastes. On the other side,
by the same axioms, we can formalize the formation/comparability of homogenous
groups of persons in a heterogeneous society, because sharing the same set of personal
conditions/preferences. In both case, indeed, the formal condition consists in the
possibility that there exists in our modeling an one-to-one correspondence from
the set of persons H to H itself, such that (the objective/subjective situation of) a
person j can be mapped onto (the objective/subjective situation of) a person k, that
is, k = ρ( j), despite originally k ̸= j.
Of course, Sen’s theory of SCF depending on a theory of justice as equity, whose
basal information space is given in terms of capabilities and functionings, poses
significant logical and then computational problems for their modeling.
The Computational Challenge of Amartya Sen’s Social Choice … 111
Sen well synthesizes the deep problems related with variable weighing in statistical
modeling of his SCF theory in the two following passages of his book ([7], pp.
358–359).
The heterogeneity of the components in the basal space, such as different functionings, points
inevitably to the need to weigh them against one another. This applies to all approaches that
respect plurality in one form or another, including the Rawlsian focus on primary goods, or
the Aristotelian focus on functionings and capabilities (which is present also in other theories
that take note of different aspects of the quality of life).
For instance, we can “close our eyes” by simply assuming that there exists something
homogenous like “the income by which everyone’s advantages can be judged and
impersonally compared, but this cannot solve the problem rather it avoids it. Indeed,
real income comparisons involve aggregation over different commodities, and in judging
comparative individual advantages there is the further problem of interpersonal comparisons,
taking note of variations of individual conditions and circumstances. It is, of course, possible
to reflect these variations in values of “adjusted income” that can be appropriately defined,
but that is only another way of stating the same problem, requiring that attention be paid to
the valuation of heterogeneous factors, though expressed in the indirect space of equivalent
incomes. Measurements in the direct space (e.g. quality of life, or capability indicators) and
those in the indirect space (e.g. equivalent incomes) would have a tight correspondence with
each other, related to the underlying values on which both the normative exercises are based.
One way or another, the issue of valuation and weighing has to be faced.
Of course, all this means that the usual mathematical and computational tools of
statistical analysis used in social and economic sciences appear inadequate, as far as
there are based on the necessity of supposing a fixed and then unique set of weights
for evaluating the variables. Indeed, the problem depends on the high variability of
the circumstances that make in principle impossible to arriving at the definition of a
unique and fixed set of weights, because of the consequent ever changing character
of the inner higher order correlations in the data set. The same problem that computer
scientists are faced when dealing with data streams. It is therefore not casual that the
main criticism to Sens SCF theory is that
it is not practically useable because of the weighing problem. In fact, however, diversity
of ingredients is only an invitation to address the problem of relative values and weights,
not an admonition to resign and go home. Indeed, when the valuation of inescapably di-
verse concerns is done implicitly as in the measurement of GDP through the use of market
prices for distinct commodities we simply withhold our responsibility to evaluate, going
instead for the mechanical use of some relative values that may have been fixed without any
relevance to normative assessment (market prices can be hugely inappropriate for ethical
evaluations, when there are externalities, asymmetric information and big inequalities in
income distribution) ([7], p. 369).
112 G. Basti et al.
On the other hand, Sen emphasizes another fundamental factor making simply ab-
surd the argument sometimes proposed that the solution for making usable and even
operational the capability approach only if it comes with a set of given weights on
different functionings in some fixed list of relevant capabilities—like for instance
in Martha C. Nussbaum’s approach (see ch. 2 of [32]). This factor is strictly re-
lated with our communication age in which valuation and weighing are in real time
systematically influenced by the public discussion and scrutiny on it
This need not, however, derail evaluation of public policy, or disable welfare-economic eval-
uations. The capability approach is entirely consistent with reliance on partial rankings and
limited agreements. The main task is to get the weights – or ranges of weights – appropriate
for the comparative judgements that can be reached through reasoning, and if the result is a
partial ranking, then we can make precisely those judgements that a partial ranking allows.
There is no obligation to feel compelled to opine on every comparative assessment that can
be proposed ([7], p. 369. Italics are ours).
It is precisely at this point that the relevance of the AI support to social choice the-
ory becomes evident. Indeed, if the problem mainly consists in getting “the weights
appropriate for the comparative judgements that can be reached through reasoning”,
because of the complexity of these judgements and the necessity to be continuously
updated often in real time, the support of AI to the human reasoning becomes essen-
tial.
On this regard, for concluding this contribution, it is essential to discuss briefly
the core of the actual discussion about the different machine learning models10 in AI
systems because their strict relationship with the problem of the statistical variable
10 By machine learning computer scientists generally intend the ability of a machine of simulat-ing
human brain ability of learning from its past experience without any further interven-tion of the
programmer. Namely, in classical or symbolic AI computational systems such as those based on M.
Minsky’s frame theory [36]—which is at the basis of contemporary “object-oriented programming”
technique—, when the machine encounters a new situation it selects from memory a given data-
structure or frame for representing a stereotyped situation, whose details can be adapted to fit
reality. Effectively, a frame is a network of nodes and relations, whose top levels are fixed because
representing what is supposed to be always true (effectively, true with the highest probability) in
a given situation, and lower levels have many terminals or slots to be filled with specific instances
or data, according to specified conditions its assignments must meet. Of course, “frames” are
programmed data-structures. Machine learning using ANN deep learning wants precisely to avoid
as much as possible this human programmer intervention in the definition of the memorized reference
patterns. In this sense, it is at the basis of the so-called non-symbolic approach to AI, also because
in the case of the “big data” this human programmer intervention is effectively impossible, because
of the complexity of the data structures involved, characterized by millions and ever billions of
parameters to be taken into account.
The Computational Challenge of Amartya Sen’s Social Choice … 113
weighing that emerges therefore as the true challenge both for the natural and the
artificial reasoning in any realm of contemporary sciences.
Effectively, the actual state of art of machine learning in AI systems is strictly
related with the different deep learning algorithms inspired by the so-called artificial
neural networks (ANN) approach in the so-called non-symbolic AI, necessary for
dealing with large sets of statistical data and of their inner and variable correlations
(see Footnote 11), as it is the case of our social choice theories.
Without entering here into the technicalities of ANN approach in machine learn-
ing, and despite the demonstrated utility of the different and more evolved learning
algorithms in solving specific problems with “big data” (see [37], for an updated
review), it is significant for us for understanding why the actual deep learning ap-
proach in AI-decisional support for human reasoning is certainly useful but not yet
sufficient in principle for addressing to the problems emphasized by Sen of a true
dynamic weighing of the variables in social choice theory—and not only in it.
Indeed, all the actual deep learning models in AI systems are based on a statistical
(not dynamic) approach to machine learning. This is essentially a modeling in which
there exists a sharp division between the “learning phase” onto some representative
sample of the database for fixing the proper set of weights representing the higher
order correlations among the variables in the dataset, and the “testing phase” for
applying fruitfully the learned weights configuration onto another representative
sample of the database, and finally, if the test is positive, onto the whole database.
Therefore, it is evident that this approach cannot work when we have to deal in
machine learning not only with huge, even though static bases of data, but with
dynamic streams of data, in which there exists an indefinite and unpredictable—
because continuously changing—number of possible correlations among an anyway
finite set of data.
That is, we are faced again with the infinity challenge in computational systems.
This means that such an approach cannot work in the actual social and economic
contexts, in which the fast information exchange among the social and economic
actors distributed all-over the world determines the sudden formation/destruction
of long-range correlations among their behaviors. In such cases, like, on the other
side, in Sen’s modeling of SCFs in which economic and social actors are directly
involved in the process to be modeled, the weights cannot be fixed in any way, as Sen
emphasized many times. In other terms, in these cases we need ANNs in continuous
learning, just it happens in biological neural networks in which no absolute separation
between the learning and testing phases occurs. The brain neurons, constituting the
hidden layers of biological NNs, indeed, not only are always oscillating, otherwise
their EEG would become flat, but the EEG traces in healthy brains are characterized
by the highest variability, without any recurrent periodicity, because a living brain is
physically a dissipative or open system in continuous exchange with its environment.
Unfortunately, for the actual statistical approach to ANNs that are based on statis-
tical mechanics and then must model ANNs as closed systems non interacting with
their environments, “continuous learning” means that the non-linear behavior of the
net becomes immediately chaotic and then in principle intractable for whichever
statistical learning algorithm such as, for instance, the well-known gradient descent
114 G. Basti et al.
algorithm that, with several updating and corrections, is the actual standard paradigm
in AI machine learning, since 80s of the last century [37–39].
According to this approach, the deep learning in the many layers of the hidden neu-
rons between the input and the output layers of the net, and whose dynamics is gov-
erned by a non-linear sigmoid function σ ,11 consists in an optimization process by the
stochastic gradient descent algorithm. That is, it consists in the calculus of the mini-
mum of an error-function (loss function), reducing recursively the distance between
the actual and the desired output of the net. This happens by “back-propagating”
recursively the error for redefining the statistical weight configuration of the connec-
tions between the input and the hidden neurons of the net, till the absolute minimum
of the error function is reached [39]. This means that the back-propagation learning
algorithm is a supervised learning algorithm since the optimization process supposes
the existence of a learning target to be reached, depending on a teacher, effectively
the human programmer.
The separation between the learning and the testing phases in this statistical stan-
dard paradigm of ANNs depends then theoretically on the necessity of defining
through the learning phase one only fixed set of neuron weights, and therefore one
only “energy landscape” for the dynamics of the net, because determining one only
“potential function” of the error to be minimized. This, therefore, requires that the
system be conceived as stable at equilibrium, and then as isolated from any other
further information exchange with its environment, otherwise the dynamics would
lose its “structural stability”. Namely, its weights would be continuously changing
11 The non-linear sigmoid function σ (x) = 1+e1 −x , and overall its immediate relative the hy-
% x −x &
perbolic tangent function tanh x = ee x −e +e−x
sinh x
= cosh x are what characterizes the learning algo-
rithms of the actual ANNs, instead of the linear step-wise 1/0 activation function of Rosenblatt’s
early “linear perceptron” [40]. This activation function is, effectively, the Heaviside function:
H (x) = ddx max {x, 0} for x ̸= 0, whose values are obviously “zero” for negative arguments and
“one” for positive arguments, so to make linear—namely, a summation of the neuron weights each
representing a variable of the dataset—the activation function of the “inner” or “deep” neurons
between the “input” and “output” and neurons. To sum up, by using the sigmoid as the activation
function, the overall neuron output is no longer 0/1, but any real value between 0 and 1, so to
allow to calculate the probability that some event (output) happens, given some conditions (input),
instead of the simple no/yes answer of the linear ANNs. The non-linearity is then fundamental
for calculating higher-order or long-range correlations among data. Indeed, mathematically, the
Taylor series expansion of tanh includes several higher order correlations, much more than sinh
and cosh alone, of which tanh is the ratio, so to answer in principle to the main criticism made by
M. Minsky and S. Papert at MIT [41] to the early linear ANNs structures, such as McCulloch’s
and Pitt’s neural net [42] and Rosenblatt’s linear perceptron [40], who first demonstrated that a
neural net is able to calculate in a desirable parallel way any Boolean logical function like a Turing
Machine. Effectively, because of the prestige of M. Minsky in US computer science community, his
criticism was able to block any further research in ANNs, from 60s to 80s of the last century. In 80s
a non-linear multilayer perceptron with a sigmoid as activation function for the hidden or “deep”
layer neurons [39], and then characterized by the gradient descent learning algorithm [38], answered
Minsky’s main criticism to early linear ANNs. However, only during the last twenty years, with the
large availability of ever more powerful processors—essentially arrays of GPUs—for reckoning
with the computational weight of this type of ANN architectures, we have the actual development
of AI-systems based on deep machine learning algorithms (see [37] for an updated synthesis).
The Computational Challenge of Amartya Sen’s Social Choice … 115
like its overall error energy landscape. Now, this non-stationary character of a dy-
namics is precisely what characterizes the notion of deterministic chaos in the theory
of non-linear dynamic systems, ANNs included [43, 44].
This finally explains because these statistical models of machine learning might
work well with big data in static bases of data, but not in dynamic streaming of data.
Moreover, the all-with-all connectivity between input and hidden neurons explains,
on the one side, why this computational architecture effectively does not answer—
theoretically—the other main objection of Minsky against ANNs. Namely that they
are not properly parallel computation architectures (for being really so, different
subsets of the hidden neurons would have to be linked only with subsets of the
input neurons and not all-with-all, so to compute different parts of the complex input
simultaneously). On the other hand, this total connectivity explains why this type
of machine learning algorithm are so computing-power (but also electric-power, of
course) consuming. Indeed, with the full connectivity, the number of the possible
combinations to be computed grows exponentially with n!, where n is the number of
neurons of the net.
This explains why this machine learning statistical approach spread effectively
only during the last twenty years, when computationally powerful GPUs became
available at chip prices on the market. Even though this approach became known to
the wide public only few years ago, effectively after the successful application by
Hinton in 2012 [45] to the ImageNet database of millions of images. An application
that earned him the front page on the New York Times. On the other hand, this “com-
putational weight” of deep learning algorithms explains why the future of machine
learning in AI is strictly related with the development of much more computationally
powerful quantum computers. It is therefore highly significant that the dynamics of
the biological neural nets of our living brains is chaotic [46]. Living brains are evi-
dently open systems in “continuous learning”, because in continuous exchange with
their environments—otherwise they would be dead brains in containers fulfilled with
formalin of some neuroanatomy lab. However, if this chaotic character make in prin-
ciple despairing any attempt of a statistical approach to NNs in continuous learning,
this opens the way to a dynamic modeling of this type of deep learning, effectively
already developed for biological NNs, but that can be applied also to ANNs. Indeed,
the connection with our previous discussion is that such a chaotic dynamics is the
demonstrated macroscopic manifestation of an underlying microscopic dynamics
of the many-body field theory of dissipative quantum systems, characterizing the
fundamental physics of the condensed matter, biological systems included [47, 48].
Moreover, the reference to dissipative QFT of many-body interacting systems we
anticipated many times in this paper as the proper physical paradigm to which also
Sen’s social theory of the many interacting groups in society must refer, in the case
of biological NNs, has another proper justification. It is, indeed, the only possible
explanation of the long-range correlation waves proper of the oscillatory behavior
of mammalian neocortex, as well as of the mammalian thalamocortical interactions
during the action-perception cycles characterizing their intentional behavior. This,
on the one hand, suggests an intriguing hypothesis about a dynamic mechanism of
multilayered long-term memory constitution (“dynamic deep learning” in quantum
116 G. Basti et al.
computing) in the mammalian brains, using the powerful QFT principle of the QV-
foliation [47, 48]. On the other hand, the operator algebra formalism underlying
this model [49] suggests the development of a new architecture of quantum optical
computer based on dissipative QFT in physics, and on Category Theory in logic
[11, 50], by which approaching from a new dynamical and no longer statistical
point of view the challenges of an unsupervised machine learning in AI [11, 51], for
dealing with data streaming and with dynamic bases of data, both characterized by
the sudden formation-dissolution of long-range correlations (field phase coherences)
among data.
Effectively, the same G. Hinton more recently stated that the future of machine
learning in AI is related with the development of unsupervised learning algorithms
in ANNs. In an interview with Steve Levin on September 15th, 2017 the interviewer
so summarized Hintons new position:
In back-propagation labels or weights are used to represent a photo or voice within a brain-
like neural layer. The weights are then adjusted and readjusted, layer by layer, until the
network can perform an intelligent function with the fewest possible errors. But Hinton
suggested that, to get to where neural networks are able to become intelligent on their
own, what is known as unsupervised learning “I suspect that means getting rid of back-
propagation. I don’t think it’s how the brain works he said. “We clearly don’t need all the
labeled data.12
13 At this point, it is significant to emphasize that in US, Todd L. Hylton and his collaborators
recently launched the Thermodynamic Computer Project (see https://knowm.org/thermodynamic-
computing/). For instance, the so-called “thermodynamic NNs” are trying to approach a type of
unsupervised solution in machine learning using the minimum free energy (in Helmholtz formaliza-
tion within the statistical mechanics approach) as a selection optimization criterion. They remain,
however, in the paradigm of the statistical interpretation of thermodynamics, also for modeling
neurons interacting with (their representation of the) environment, using an ingenious, but anthro-
pomorphic logic of “ question/answer” [53]. In this way, the dynamics of such a “thermodynamic
neural network” is a sort of stroboscopic approximation (i.e., limited to one phase or to a small num-
ber of close phases) of what in QFT is a continuous process of phase transitions (macroscopically, a
chaotic dynamics) and then, when implemented computationally, a universal thermodynamic LTS.
In this limited sense, this class of ANNs is straddling (2) and (3) types above of unsupervised
machine learning algorithms.
118 G. Basti et al.
References
1. Hansson, S.O., Hendricks, V.F. (eds.): Introduction to Formal Philosophy. Springer, Berlin &
New York (2018)
2. Endriss, U.: Logic and social choice theory. In: Gupta, A., Van Benthem, J. (eds.) Logic and
Philosophy Today, pp. 333–377. College Pubblications, London (2011)
3. Bergson, A.: A reformulation of certain aspects of welfare economics. Q. J. Econ. 52(2),
310–334 (1938)
4. Samuelson, P.A.: Foundations of Economic Analysis. Harvard University Press, Cambridge,
MA (1947)
5. Arrow, K.J.: Social Choice and Individual Values, 2nd edn. Yale University Press, New Haven
& London (1963)
6. Sen, A.K.: Collective Choice and Social Welfare. Elsevier, Amsterdam (1979)
7. Sen, A.K.: Collective Choice and Social Welfare, Expanded edn. Penguin Ltd., Kindle Edition,
London (2017)
8. Sen, A.K.: The impossibility of a Paretian liberal. J. Polit. Econ. 78(1), 152–157 (1970)
9. Rawls, J.: A Theory of Justice. Oxford University Press, Oxford, UK (1971)
10. Stiglitz, J.E., Sen, A.K., Fitoussi, J.-P.: Mis-measuring our lives. Why GDP doesn’t add up. The
report by the Commission on the Measurement of Economic Performance and Social Progress.
With a foreword by Nicolas Sarkozy. New Press, New York (2010)
11. Basti, G., Capolupo, A., Vitiello, G.: Quantum field theory and coalgebraic logic in theoretical
computer science. Prog. Biophys. Mol. Biol. 130, 39–52 (2017)
12. Sen, A.K.: The Idea of Justice. Penguin Book, London (2010)
13. Sen, A.K.: Development as Freedom. Anchor Books, New York (2010)
14. Tirole, J.: Economics for the Common Good. Princeton University Press, Princeton, NJ (2017)
15. Basti, G.: The post-modern transcendental of language in science and philosophy. In: Delic,
Z. (ed.) Epistemology and Transformation of Knowledge in a Global Age, pp. 35–62. InTech,
London (2017)
16. Mulgan, G.: Big Mind. How Collective Intelligence Can Change Our World. Princeton Uni-
versity Press, Princeton, NJ (2018)
17. de Condorcet, M.: Essai sur lApplication de lAnalyse à la Probabilité des Decisions Rendues
à la Pluralité des Voix. L’Imprimerie Royale, Paris (1785)
18. Kirman, A., Sonderman, D.: Arrow’s theorem, many agents, and invisible dictators. J. Econ.
Theory 5(2), 267–277 (1972)
19. Mihara, H.R.: Arrow’s theorem and Turing computability. Econ. Theory 10(2), 257–276 (1997)
20. Sen, A.K.: The informational basis of social choice. In: Maskin, E., Sen, A.K. (eds.) The Arrow
Impossibility Theorem, pp. 67–100. Columbia University Press, New York (2014)
21. Rawls, J.: Jusitice as fairness. Philos. Rev. 67(2), 164–194 (1958)
22. Rawls, J.: Justice as fairness: political not metaphysical. Philos. Public Aff. 14, 223–252 (1985)
23. Rawls, J.: Political Liberalism. Columbia University Press, New York (1993)
24. Rawls, J.: In: Kelly, E. (ed.) Justice as Fairness. A Restatement. The Belknap Press of Harvard
University Press, Cambridge, MA & London, UK (2001)
25. Kant, I.: Grundlegung zur Metaphysik der Sitten. Johann Friedrich Hartnock, Riga (1785)
26. Kant, I.: Fundamental Principles of the Metaphysics of Ethics (trans: Abbott, T.K.). Longmans,
London (1907)
The Computational Challenge of Amartya Sen’s Social Choice … 119
27. Sidgwick, H.: The Methods of Ethics, vol. 3. McMillan, London (1874)
28. Hare, R.M.: Freedom and Reason. Oxford University Press, Oxford, UK (1963)
29. Hare, R.M.: The Language of Morals. Oxford University Press, Oxford, UK (1963)
30. Rawls, J.: The priority of rights and the idea of good. Philos. Public Aff. 17(4), 251–276 (1988)
31. Suppes, P.: Some formal models of grading principles. Synthese 16(3–4), 284–306 (1966)
32. Nussbaum, M.C.: Human functioning and social justice: in defense of Aristotelian essentialism.
Polit. Theory 20(2), 202–246 (1992)
33. Nussbaum, M.C.: Creating Capabilities. Harvard University Press, Cambridge, MA (2011)
34. Alkire, S.: Valuing Freedoms: Sens Capability Approach and Poverty Reduction. Oxford Uni-
versity Press, New York (2002)
35. Smith, A., : In: Sen, A.K. (ed.) The Theory of Moral Sentiments, 1776. Penguin, London (2009)
36. Minsky, M.: Frame theory. In: Johnson-Laird, P.N., Wason, P.C. (eds.) Thinking: Reasings in
Cognitive Science, pp. 355–376. Cambridge University Press, Cambridge, MA (1977)
37. Nwankpa, C.E., Ijomah, W., Gachagan A., Marshall, S.: Activation Functions: Comparison
of Trends in Practice and Research for Deep Learning. https://arXiv.org/arXiv:1811.03378v1
(2018, 8 Nov). Accessed 30 Jun 2019
38. Werbos, P.: Beyond regression: new tools for prediction and analysis in the behavioral sciences.
Ph.D. thesis. Harvard University Press, Boston, MA (1974)
39. Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning representations by back-propagating
errors. Nature 323(6088), 533–536 (1986)
40. Rosenblatt, F.: Principles of Neurodynamics. Perceptrons and the Theory of Brain Mechanisms.
Cornell University Press, Buffalo, NY (1961)
41. Minsky, M., Papert, S.: Perceptrons. An Introduction to Computational Geometry, 2nd edn.
MIT Press, Cambridge, MA (1987)
42. McCulloch, W.S., Pitts, W.H.: A logical calculus of ideas immanent in nervous activity. Bull.
Math. Biophys. 5(4), 115–133 (1943)
43. Hilborn, R.: Chaos and Non-linear Dynamics. Oxford University Press, Oxford (1994)
44. Viana, M.: What’s new on Lorenz strange attractors. Math. Intell. 22(3), 6–19 (2000)
45. Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional
neural networks. In: Proceedings of the 25th International Conference on Neural Information
Processing Systems, vol. 1, pp. 1097–1105. Curran Associates Inc., Lake Tahoe, Nevada (2012)
46. Freeman, W.J.: How Brains Make Up Their Minds. Columbia University Press, New York
(2001)
47. Freeman, W.J., Vitiello, G.: Dissipation and spontaneous symmetry breaking in brain dynamics.
J. Phys. A: Math. Theor. 41(30), 304042 (2008)
48. Freeman, W., Vitiello, G.: Matter and mind are entangled in two streams of images guiding
behavior and informing the subject through awareness. Mind Matter 14(1), 7–24 (2016)
49. Blasone, M., Jizba, P., Vitiello, G.: Quantum field theory and its macroscopic manifestations.
Ordered Patterns and Topological Defects. Imperial College Press, London, Boson Condensa-
tion (2011)
50. Rasetti, M., Marletto, P.: Quantum physics, topology, formal languages, computation: a cate-
gorial view–an homage to David Hilbert. Perspect. Sci. 22(1), 98–114 (2014)
51. Rasetti, M., Merelli, E.: Topological field theory of data: mining data beyond complex networks.
In: Contucci, P., Giardinà, C. (eds.) Advances in Disordered Systems, Random Processes and
Some Applications, pp. 1–42. Cambridge University Press, Cambridge, UK (2017)
52. Marcus, G.: Deep learning: a critical appraisal. https://arxiv.org/ftp/arxiv/papers/1801/1801.
00631.pdf. Accessed 15 Jul 2019
53. Fry, R.L.: Physical intelligence and thermodynamic computing. Entropy 19, 107 (2017)
54. Magnani, L.: AlphaGo, locked strategies, and eco-cognitive openness. Philosophies 4, 8–24
(2019)
55. Alkire, S., Foster, J.E., Seth, S., Santos, M.E., Roche, J.M., Ballon, P.: Multidimensional Poverty
Measurement and Analysis. Oxford University Press, Oxford, UK (2015)
56. Pattanaik, P.K.: Social choice and voting. In: Hansson, S.O., Hendricks, V.F. (eds.) Introduction
to Formal Philosophy, pp. 839–853. Springer, Berlin & New York (2018)
Could an Electronic Person Exist?
Robots and Personal Responsibility
The authors have equally contributed to the ideas and content of this article. Claudia Stancati is
responsible for Sects. 2 and 3. Giusy Gallo is responsible for Sects. 1, 4 and 5. The first version
of this text has been presented at AISB 2017 Convention and it has been published in AISB 2017
Proceedings.
1 Introduction
The term “electronic person” appeared for the first time in the European Parliament
resolution with recommendations to the Commission on Civil Law Rules on Robotics
in 2017. What does “electronic person” mean and why it is relevant to the ethical
and legal discussion on the relationship between (human) persons and robots to coin
that term?
The urgent need of a political consideration marks the necessity to find an answer
to the question about the features human beings and artificial beings should share.
While it is a simple task to enucleate the features of artificial beings, we maintain the
philosophical urgency to begin to think the concept of person again. For this reason,
the starting point is the meaning of the term persona in Roman law and the way the
concept has changed in Western law, including the various notions of legal person.
The main theme in question is personal responsibility, not only for human actions
but also for robotic ones, since robots—for example, social robots—have started to
share daily actions with human beings.
The European report mentioned above focused the attention on the developments
in robotics paying attention to the preservation of human safety and dignity starting
from the fact that juridical systems still today are not adequate to face situations in
which a robot is “guilty”.
In “Alienable rights” (1995), Minsky [1] plays out a dialogue between two aliens
speaking of the rights to be granted to the human beings. He wonders if we need to
allow that the world is ‘pollute’ with ‘redundant intelligences’.
We are ever closer to the moment in which we must seriously respond to the crucial
question on Artificial Intelligence: which features human beings and artificial beings
share? Recently, Floridi has expressed his concern about this point: “the issue is
not to decide whether the robots, as autonomous agents which are able to learn, are
such a kind of person or not, but we should understand that the question is badly
formulated” [2].
Actually, the question is particularly complex not only because we are considering
the human being close to the artificial one, but because the term “person” is everything
except that defined in an univocal way. Our aim is to look for some hints in the facets
of the debate on this notion in order to define the ontology of these entities which in
the long run could make decisions or act autonomously. This is the symmetrical case
to the one which claims for the concepts of the extended mind and the electronic
body, which have been made actual by the new technologies and have reshaped the
borders of the old philosophical debate of the mind-body problem [3].
The semantic adventure of the term ‘person’ led the “mask” to become what is
essential: the persona ficta has become what is most essential of a subject.
Could an Electronic Person Exist? … 123
The concept of the person has now entered the common language and it defines
every individual in its unique peculiarity. Philosophically speaking, the concept of the
person has a long history which also crosses the religious tradition and particularly
Christianity; today this concept has landed at the center of the discussion in bioethics,
in cognitive sciences, passing through sociology, from Marcel Mauss onwards [4].
In this story, ‘individual’ and ‘person’ have been considered different concepts but
only in relatively recent times they have been opposed one another: the first one has
been managed such as a naturalistic category and the second one such as an ethical
concept or a notion which can define a center of action.
The concept of the person, superimposed on the problem of subjectivity and
strictly tied to the theme of the values and their intention acceptance, is not clear
in its ethical and epistemological implications. We would mention the differences
among the notion of person forged by Peter Singer or Martha Nussbaum and the role
played by the ability to act, the biological nature and biography in defining those
ideas.
The different notions of the person refer to different ontological levels and because
of them we would like to consider here a concept of person drawn in legal field in
order to verify whether this one, once opportunely adjusted, could be the right tool
to define the status of robots and androids.
The crucial point of the question is to recollect a definition of the person suitable to
the electronic personality with reference to the attempt of replicating human behavior,
emotion and way of thinking [5], facing the issue of the imputation of liability, since
the future could lead us in a dimension of co-working with robots.
In Roman law, the term ‘persona’ is equivalent to homo; then it evolves to an abstract
meaning to designate who has a juridical role. This is a tendency which lead to the
concept of the persona ficta or rapraesentata like the “fictum positum pro vero” [6,
D-48, ii9, 16, 10].
A very long process of doctrinal elaboration, which is parallel to the history of
the Western law, particularly but not only the Roman juridical tradition,1 lead to
dissociate the concept of the juridical person from the legal entity, the imputation
of liability from the conferring of the legal personality. The last concept does not
necessarily refer anymore to the man but it’s linked to the powers and the interests
of the man, which are concentrated in accordance with predicative structures and
normative schemes, or it refers to complex subjects to whom have been conferred
features suitable to action both in the field of private law and public law.
This sort of persons is distinguished from natural elements, or rather the juridical
person arises as opposed to the natural element as persona ficta, both with reference to
1 The term ‘person’ is not present in the Code Napoléon, but one can find it in the common law, as
it shows the Judicial Dictionary by Stroud [18].
124 C. Stancati and G. Gallo
a single individual and to a group of persons. Law breaks down bodies and subjects to
compose persons namely masks which could personam sustinere o gerere, personam
vicem gerere. The same individual can be invested with different personae; many
subjects can represent one unique persona, whose unity is not superimposed to the
unity of a physical or psychological subject. Moreover, in some subjects there can be
the rights in the absence of a dead physical or unborn person, and on the other hand
the concept of physical person renames, for the purposes of the law, the biological
bodies (e.g. let’s consider bioethics).
The notion of legal person adapts to the definition of the artificial subjects which
are allowed on a merely functional level and, then, to every artificial individual.
In his Reine Rechtslhere Hans Kelsen maintains that: “one might consider (…)
not individuals such as, but only their actions and omissions set out in legal rules”
[7] so that the physical person “is not a man but the personified unity of legal rules
that give rights and duties to the same man. It is not a natural reality but a legal
construction” in order to describe “legally relevant facts” [7].2
This way Kelsen, who obviously in the time he wrote can apply law only to
human beings, makes a legal person without biological constraints and he rejects the
anthropomorphic view of the concept of the person.
Both the physical and the legal person are social realities such as the same law
allow us to understand that whether “the law does not create the persons” [7],3
society can regulate a new reality described by technology using this kind of concept
of person.
Uberto Scarpelli, who has introduced the analytical philosophy in Italy, reached
the thought of Kelsen; according to Scarpelli, the condition of use of the concept of
the legal person “is in the normative forecast of a plurality of actual and possible
facts, mere facts or actions, a connection among them, which consists in the reference
to a same legal system and the attribution of a juridical relevance to such facts in
terms of relationships intervening among them” [8].4
We can use this “limitedly technical” [9]5 concept to regulate new experiences
and new realities the way it arose to answer to the new situations in which “individual
subjects find multiplied and widened their limited life in a lot of new realities” [9].6
In Anglo-saxon legal literature we also find similar positions. According to Radin,
legal personality is “a convenient symbol, a short-hand expression by which a great
many factors can be designated by a single word or a single phrase” [10, p. 625, 643]
and the concept of person is “a means of making reference to complex situations easy,
and is only an inexperienced person that is deceived by it, or fancies that calling the
2 Translation from the Italian: H. Kelsen, Reine Rechtslhere, 1960, trad. it. La dottrina pura del
diritto, Torino, Einaudi, 1952, p. 193.
3 Translation from the Italian: La dottrina pura del diritto, Torino, Einaudi, 1952, p. 198.
4 Translation from the Italian: U. Scarpelli, Contributo alla semantica del linguaggio normativo,
series a, b, c, …, n, by the name of x, add a new item to the series” [10, 665]. Wolff
speaks about abbreviation and a formula of fiction, starting from the natural extra-
legal concept of personality and entities considered such as persons despite the lack
of human dignity [11, 494, 505].
Rules are a fundamental mechanism of connection between the social and the
mental by the means of the so-called “multiagent structure” that expresses itself in
the plurality of the recipient of the majority of legal rules. This structure allows
an education which falls beyond the one imagined by Kant, for whom from the
concept of legal entity emerges only and necessarily the relationship between free
and equal people and for this reason, according to him, legality cannot connote neither
our relationship with animals nor the one with God or, we can say today, with the
androids. However, the process of de-psychologizing, abstraction, de-naturalization
that features the rule of laws allows to them to refer to different models of social
organization and it shows how they are semantically multipotent thanks to the ability
to use natural elements in an artificial context, in which sometimes by the means of
the rule “x counts as y in the context c” [12] it occurs the transition from the physical
object to the social one.
According to Nèkàm the feature that makes a legal entity is the fact that the
collectivity considers it as a unity which has interests that need or deserve social
protection: “and it is completely irrelevant that such entity exists as an objective
reality, that has its will, and that it possesses or less a natural or artificial, imagined
or proven personality” [13, 21]. Following this perspective, we can think to look for
the theoretical grounds to reshape the new legal challenges that Artificial Intelligence
begins to set.
The theoretical issue that we are facing concerns the chance that artificial systems,
which could at least show a certain degree of human agency in case of work sharing
between robots and humans, would benefit of their own juridical status.
Starting from the first studies, the research in the field of Artificial Intelligence
focused on some classical philosophical issues: learning, perception and the rela-
tionship between mind and body. Despite this beginning, it is not at all obvious that
the current developments in Artificial Intelligence, particularly in robotics, claim for
a philosophical approach to the newborn (and future) relationships between human
beings and robots.
The continuous online life we are experiencing is shaping our way of living in the
information society claiming—as Floridi [14] underlines—for a new ethical frame-
work, in which we should ward off the anthropocentric view over robots allowing
to accept them as co-workers, including the huge amount of legal, social and ethical
issues at the stake.
126 C. Stancati and G. Gallo
7 Link:
http://www.carscoops.com/2016/12/teslas-autopilot-anticipates-huge-crash.html.
8 Link: http://www.europarl.europa.eu/committees/en/juri/subject-files.html?id=
20170202CDT01121.
Could an Electronic Person Exist? … 127
First of all, we should consider the issue about the protection of the persons who
share a daily experience with a robot in order to avoid potential damages to them. It is
the case of social robots. A social robot is an autonomous or semi-autonomous robot
that is involved in some programmed and scheduled practices which are structured
taking into account the role they play in interactions with human beings and/or
other social robots. Care robots and butler robots are two kinds of social robots. The
care robot is programmed to help elderly persons in their everyday life while the
butler robot acts like a servant. One critical juncture is the chance of disruptions
and malfunctions of the social robot and eventual damages to the persons he has to
take care of. Who will be responsible for those damages? The mentioned European
motion is an attempt to find an answer, facing legal, ethical and social issues. Floridi
[14] offered us a clear analysis of the motion, stating that its most relevant features
are the issues of unemployment and liability. We could add to these issues that
developments in robotics and artificial intelligence can and should be designed in such a way
that they preserve the dignity, autonomy and self-determination of the individual, especially
in the fields of human care and companionship, and in the context of medical appliances,
‘repairing’ or enhancing human beings [18, 5].
This idea covers the chance that developments in robotics and AI could be helpful
to people but their application need to be constantly checked since one cannot ignore
the users are not simply individuals but persons which sometimes have a reduced
autonomy but they have the right to preserve their dignity.
Back to the two features stressed by Floridi, we should recall the example of the
driverless or autonomous car. Let’s say, for example, that within a couple of years we
could travel with a Google or Tesla car. It makes sense that taxy service companies
might fire their drivers, just like bus and train companies since there will be no more
need to pay people for a job they will not do anymore. So that there might be a lot
of unemployers. If this scenario should happen, this means that the social relapse
will be dramatic, so that today there is already the need to think about an economic
support for people who will lose their job because of robots and algorithms. Then,
companies will save money and will reduce complaints, but the social systems of
some Nations will be troubled by this new developing situation. The only solution
is to tax robots, which means to tax the companies which replace a lot of human
workers with robots.
Getting back to the matter of responsibility, what if an autonomous car crashes
against another car drawn by a human being? Who is legally liable for the accident?
Apparently, nobody or, at least, the company who has produced the car if there is a
malfunction or the owner who made a mistake in running the system. However, the
main question concerns whether our legal systems are ready to face these situations
and what are the consequences on the notion of the person.
The previous paragraph showed that there is not an univocal notion of the person.
Since concepts such as rights, responsibility, ethics, privacy have been ascribed to
persons and not to material things such as robots, how to shape the concept of the
person?
128 C. Stancati and G. Gallo
The motion proposed to the European Parliament seems to solve the problem
creating the notion of ‘electronic person’, a new legal status to be given to autonomous
robots:
creating a specific legal status for robots in the long run, so that at least the most sophisticated
autonomous robots could be established as having the status of electronic persons responsible
for making good any damage they may cause, and possibly applying electronic personality
to cases where robots make autonomous decisions or otherwise interact with third parties
independently [18, 19].
Although the motion highlights the relevance of the attribution of the juridical
status, it does not clarify the contexts of its application. Unfortunately, the term
‘electronic person’ appears only once in the whole motion (the last version is a
document of 68 pages). This recommendation to the European Parliament is not
deeply explained or investigated.
It seems that the legal commission, taking into account the difficulties in inves-
tigating the notion of the person as it is clear in the history of the philosophy of
law, has proposed the notion of ‘electronic person’ in order to begin to think to the
consequences of robots on human beings.
The purpose of the commission which uses the term ‘electronic person’ seems a
mistake: since an ‘electronic person’ is a robot, then it is an artifact, which is some-
times, in particular, an Information Technology artifact, and it has to be considered
an artificial system. Unfortunately, this artificial system could be invested of rights,
duties, legal liability. If this should happen, then it will make sense to start think-
ing about robots such as Isaac Asimov’s Bicentennial Man, ascribing to robots—in
future—also emotions, according to the most wide meaning this word can cover.
The anthropomorphic view risks to make us think the problem of the relation-
ship between robots and persons—the persons we are—following the path which
sometimes we run thinking about the relationship between human beings and ani-
mals. However, there is a big question: animals are not artificial systems, but robots’
behavior is always governed by rules imposed by their creators.
We can say, agreeing with Floridi, that the relevance of the issue needs to be
considered changing the point of view. So then, we need to reshape the notion of
electronic person, giving sense to the terms electronic and person in a philosophical
perspective. In particular, in the previous paragraphs, we showed that the philosophy
of law of Scarpelli, Kelsen and Nèkàm, for example, ward off the biological constrain
which generally is ascribed to the person, trying to describe a new kind of society in
which there is a widespread co-working with robots and androids.
5 Conclusion
The remarks on the term “electronic person” are not conclusive by the very nature of
the whole issue. First of all, there is no chance to strictly define the electronic person
without living room for future reviews by reason of the developments in robotics
Could an Electronic Person Exist? … 129
and Artificial Intelligence. Paragraph 2 showed that there is not a univocal concept
of person and legal person. Ancillary concepts such as responsibility and ethics get
involved also in the issue of electronic personality that is an artificial personality.
How to make consistent a system of qualities, generally ascribed to human beings,
recently at stake because they could be also ascribed to artificial beings? One could
simplify the question such a new sort of opposition between natural and artificial.
On the contrary, we are facing the risk of an anthropomorphic view, while we need a
new ethical and philosophical perspective which follows over the year the prospective
relationship between robots and persons.
References
1. Minsky, M.: Alienable rights. In: Android Epistemology, pp. 307–312. MIT Press, Cambridge,
MA (1995)
2. Floridi, L.: L’ultima legge della robotica. La Repubblica (2017)
3. Gallo, G., Stancati, C.: Where is the mind? The extended mind reloaded. In: Proceedings of
Annual Convention of the Society for the Study of Artificial Intelligence and the Simulation
of Behavior (AISB 2016), pp. 242–249. Curran Associates, New York (2017)
4. Carruthers, M., et al. (eds.): The Category of the Person. Anthropology, Philosophy, History.
Cambridge University Press, Cambridge (1985)
5. Minsky, M.: The Emotion Machine. Commonsense Thinking, Artificial Intelligence and the
Future of the Human Mind. Simon & Schuster, New York (2006)
6. Bartolo di Sassoferrato: Commentarium in Digestum Novum, c. D. 48, ii9, 16, 10
7. Kelsen, H.: Reine Rechtslhere. Reine Rechtslehre. F. Deuticke, Wien (1934)
8. Scarpelli, U.: Contributo alla semantica del linguaggio normativo. Giuffrè, Milano (1959)
9. Capograssi, G.: Il problema della scienza del diritto, new edition. Giuffrè, Milano (1962)
10. Radin, M.: The endless problem of corporate personality. Columbia Law Rev. 32, 643–667
(1932)
11. Wolff, M.: On the nature of legal persons. Law Q. Rev. 54(4), 494–521 (1938)
12. Searle, J.: Mind, Language and Society. Basic Books, New York (1998)
13. Nékàm, A.: The Personality Conception of the Legal Entity. Harvard University Press,
Cambridge, MA (1938)
14. Floridi, L.: Roman law offers a better guide to robot right than sci-fi. Financial Times (2017)
15. Lovink. G.: Network Without a Cause. Polity Press, London (2012)
16. Floridi, L.: Robots, jobs, taxes and responsibilities. Philos. Technol. (2017). https://doi.org/10.
1007/s13347-017-0257-3
17. Ruffolo, U.: Per i fondamenti di un diritto della robotica self-learning: dalla machinery pro-
duttiva all’auto driverless: verso una “responsabilità da algoritmo”? In: Ruffolo, U. (ed.)
Intelligenza Artificiale e responsabilità, pp. 1–30. Giuffrè, Milano (2017)
18. Committee on Legal Affairs (Rapporteur Mady Delvaux): European Parliament resolution with
recommendations to the Commission on Civil Law Rules on Robotics. http://www.europarl.
europa.eu/committees/en/juri/subject-files.html?id=20170202CDT01121
19. Stroud, F.: The Judicial Dictionary of Words and Phrases Judicially Interpreted. Sweet &
Maxwell, London (1890)
Habits, Routines and Rituals
Cognitive Dynamics of Research
Routines: Case Study of MicroRNA
Paweł Kawalec
1 Introduction
The idea to systematically investigate science as a social practice was first intro-
duced in the 1920s and 1930s by the founders of “the science of science” move-
ment [26] Znaniecki [57], Ossowska and Ossowskis [41, 42] and independently by
Bernal [8]. Maria and Stanisław Ossowskis elaborated on the former work of Flo-
rian Znaniecki and identified three core disciplines that constitute scientific study of
science: epistemology and philosophy of science, psychology of scientific creativity
and anthropology or sociology of science. They argued that understanding science
as a social practice is necessary in order to grasp the dynamics of science, which can-
not be uncovered by studying its “immanent factors” alone [42, p. 75]. The overall
argument presented by Ossowskis was meant to support the idea of an autonomous
discipline of “the science of science” and, in consequence, give credit to the plans
of launching a dedicated research institute in the late 1930s. These plans, however,
P. Kawalec (B)
Faculty of Philosophy, John Paul II Catholic University of Lublin, Lublin, Poland
e-mail: pawel.kawalec@kul.pl
did not materialize because of the outbreak of WWII and the post-war communist
regime in Poland.
An alternative proposal was advanced by John Bernal in his book The Social
Function of Science (1939). It was, at least partly, motivated by “disillusion” due
to negative effects of science development, such as unemployment, inequality or
new weapons threatening “personal security”. Moreover, as Bernal observed, there
was “a strange coincidence, the disturbing events of the Great War, the Russian
revolution, the economic crises, the rise of Fascism, and the preparation for newer
and more terrible wars have been paralleled inside the field of science by the greatest
changes in theory”, such as emergence of mathematical logic, relativity and quantum
mechanics, biochemistry and genetics [8, p. 2]. So, he set out to elaborate on “the
fact that science is both affecting and being affected by the social changes of our
time”, emphasizing that “this is a social and economic rather than a philosophical
inquiry” [8, p. 9]. For Bernal, then, overall “Scientific research and teaching are in
fact small but critically important sections of industrial production” [8, p. 10].
Admittedly, he was first to provide ample, but necessarily incomplete, evidence
establishing the link between the economic growth and intensity of R&D. In fact, he
significantly contributed to the UK almost doubled increase in R&D public expendi-
ture of the post-war period. Bernal’s unwillingness to pursue the underlying mech-
anism of the effectiveness of economic regularities related to R&D spending was
inherited by the post-war mainstream economists like Paul Samuelson, Kenneth
Arrow, Richard Nelson and Robert Solow.
Slightly later, in the early 1950s, Derek de Solla Price initiated the turn of “the
tools of science on science itself” (1951). He explicitly motivated it by an analogy
with physics: “The method to be used is similar to that of thermodynamics, in which is
discussed the behavior of a gas under various conditions of temperature and pressure.
One […] considers only an average of the total assemblage in which some molecules
are faster than others, and in which they are spaced out randomly and moving in
different directions. On the basis of such an impersonal average, useful things can
be said about the behavior of the gas as a whole, and it is in this way that I want to
discuss the analysis of science as a whole” [17, p. xiv].
However, it was only after WWII that this idea really took off. Although the
main perspectives, such as philosophy and history, economics, scientometrics and
sociology significantly differed, they seem to have converged on a common topic of
the growth of knowledge.1
In the remainder of this paper I focus on the philosophical and descriptive per-
spective on the cognitive aspect of this dynamics, leaving aside its social aspect.
Important modern contributors, such as F. Bacon or Descartes, predate the contem-
porary philosophers of science in their search for “New Organon”, “the logic of
induction” or “the logic of discovery”. During the early decades of the 20th cen-
tury the analysis of scientific discoveries was predominantly “logical” [44, p. 3],
which reflected the then recent canonical formulation of the mathematical logic.
Nonetheless, already forty years ago the narrowness of this approach was forcefully
questioned [21, p. 120] by, among others—what was later dubbed as—“the friends
of discovery” [40]. In effect, it paved the way for a new research on science dynamics
that has given rise to a plethora of different approaches.
Following [22] I overview the main threads in these accomplishments and char-
acterize the contribution of the present paper within this thriving research field. The
deductive approach develops the traditional approach of logical empiricists, in par-
ticular Popper’s. The main claim is that while science progresses by constructing
“hypotheses, or systems of theories” and testing “them against experience by obser-
vation and experiment”, it is the task of deductive logic to reconstruct “this procedure”
and “analyse the method of the empirical sciences” [44, p. 3]. The obvious objec-
tion is that deductive logic is non-ampliative, so deductions can merely make more
explicit the information that was already presumed in the premises. Emiliano Ippoliti
debunks the main counter-arguments and demonstrates that the deductivist approach
cannot account for generation of hypotheses as “the crucial step in discovery” [23].
The cognitive approach provides an alternative perspective on scientific discovery.
It was initiated by Simon [47] and characteristically claims that scientific discoveries
as well as “creative products of all sorts are brought about by our ordinary cogni-
tive processes, such as those involved in our day-to-day problem-solving activities”
[55, p. xi]. Further, it is claimed that these cognitive activities are rule-governed
and transparent to cognitive agents themselves, so, in principle, it is possible to
implement them in any kind of information-processing systems, such as computers
[31]. Nevertheless, three objections to this approach still seem to hold. Firstly, the
implementations, such as the most famous software BACON, are not capable of per-
forming the crucial step in the generation of a new hypothesis, namely identification
of relevant variables and data. Secondly, by analogy to the deductivist approach, they
merely reconstruct the historical processes and, at best, discover a new solution to
a problem that has already been recognized and solved independently. Thirdly, this
cognitive system is incapable of generating new conceptualizations that would drive
subsequent formulations of problems and their solutions.
Evolutionary approach provides yet another way to conceive of scientific discov-
ery. In its traditional formulation [14] it mimics biological processes in claiming blind
generation of hypotheses that are subsequently subsumed to processes of selective
retention during acquisition of new data or new theoretical models. One of the objec-
tions to this approach is that the process of scientific discovery is not intentional and
is not guided by methods or rules. The other objection is that it is, in principle, not
possible to articulate the aim of science as an endpoint of an evolutionary process.
While my approach presented in this paper takes the criticisms and limitations
of the three aforementioned approaches for granted, it builds on the evolutionary
approach. In my earlier publications [25, 27, 28] I elaborated heuristics-driven vari-
ation + institutional selective retention (HDV + SR) model of two-fold cognitive-
social dynamics of science. It claims, roughly, that the hypotheses are generated as
projections of the previously established body of scientific knowledge that are further
explored in order to ground them empirically [53]. If successful, the process may ini-
tiate a growing institutional set-up that promotes diffusion of the hypotheses among
136 P. Kawalec
the members of scientific community and may bring about a new research routine
as a recognizable social practice of research sharing a symbolic representation.
HDV + SR model has several characteristics that stand out from the traditional BV
+ SR model. First, it takes into account a more complex ontology of scientific change,
including not only ideational generations and re-combinations, but also an associated
social ontology [19]. Second, it takes the growth of knowledge to constitute an
intentional process that is rule-governed. Third, it propounds a processualist approach
to the growth of knowledge [39]. The traditional model of “scientific discovery”
presumes that there are some special “turning points” that orient the development
of the whole research field. In contrast, I favor the perspective that takes the growth
of knowledge to be a continuous process which, of course, exhibits variations in its
dynamics that are due to interaction of different underlying mechanisms and shock
impulses. The present paper uses the case of one of the major recent breakthroughs
in molecular biology, namely microRNA, to elaborate in greater detail on the latter
point and to strengthen the processualist aspect of HDV + SR model by identification
and examination of the aforementioned mechanisms.
It has to be remarked that the terminology used to discuss the issues related to
scientific discovery and its various kinds, such as “breakthrough”, “novel” or “trans-
formatory”, is far from clear. The use of the very term “scientific discovery” has
ranged from meaning a single outcome to referring to the whole procedure of inquiry
[46]: “In the narrowest sense, the term ‘discovery’ refers to the purported ‘eureka
moment’ of having a new insight. In the broadest sense, ‘discovery’ is a synonym
for ‘successful scientific endeavor’ tout court”. The latter use, however, has steadily
disappeared from modern philosophy of science with its gradual divergence from
scientific research.
Apparently, the term “scientific discovery” still connotes the kind of approach in
philosophy of science which I would call here “particularist” as it tends to focus on
pinning down particular moments of radical change in science. Most often they are
understood as manifested by key publications that are indicated as turning points in
the history of science and are recognized by prestigious scientific awards. Because
I would like to avoid the particularist presumption here, I am inclined to use an
alternative and more processualist-oriented terminology that allows me to carve out
more complex wholes within the process of inquiry.2 It appears that research process
is the obvious candidate for such a unit, granted a processualist approach, as it
delivers the outcomes which are then disseminated in publications and it is relatively
2I adopt here the terminological distinctions developed in the debate between proponents of
particularist and processualist anthropology (see esp. [10, 11]).
Cognitive Dynamics of Research Routines: Case Study of MicroRNA 137
well discriminable from other research processes by its unique research question, the
adopted methodology, its specific research design, the collected data and results.3
As argued in my earlier publications, my preferred way to approach the dynamics
of research processes is in terms of research routines. By “research routine” I mean
a repeated and recognizable pattern of research practices of a scientific community
which share a symbolic representation, such as a model or theory. Research routine
presumably constitutes the main source of the heuristics-driven variance for each time
it is used in a particular research process it is enacted in a modified way due to specific
research question and research design. In effect, it is accordingly used to project new
hypotheses. Nevertheless, in most research processes research routines remain stable
and the respective changes do not affect the core of the symbolic representation, but
only expand it as they become more specialized and more widely applicable. Only
occasionally some components of the symbolic representation undergo a signifi-
cant modification and eventually a new research routine emerges. However, it is not
accomplished in a single research process, but lasts for an extended period of time
and remains continuous between the old and the emergent routine as exemplified by
the phases in the development of microRNAs [28].4
As mentioned in Sect. 1, research routines result from the twofold dynamics:
cognitive and social. The cognitive progress is evidenced by the appearance of new
concepts, new distinctions and links between them, new areas of application, etc.
The social dynamics is evidenced by the growing institutionalization of a given
research routine, such as getting from an informal group of researchers through
acquiring large-scale research projects, establishing standards (e.g. terminology,
databases, protocols), setting up dedicated graduate educational programs and pub-
lication of monographs, and new formal institutions (e.g. chairs, departments, insti-
tutes, research laboratories etc.), up to formation of subdisciplines, specialized con-
ferences, learned societies and journals. Although these two dynamics are integrated
in the growth of research routines, for the purpose of the present paper I try to single
out the cognitive dynamics and determine its underlying mechanisms.
3 A general and forceful case for the processualist turn in the philosophy of life sciences has recently
issue of how new knowledge expands the existing body of knowledge, such as new
classifications (e.g. subcategories of microRNAs) or new relationships between the
existing categories (e.g. microRNAs as tumor suppressors). And, finally, utility mea-
sures the extent to which the new idea is useful. Simonton set the scope of values for
each of the three variables in analogy with probabilistic measure ranging between
0 and 1. The product of the three variables constitutes the measure of scientific
creativity.
As argued elsewhere [25, 27, p. 133], for epistemological reasons I do not follow
Simonton’s procedure for parametrization of his creativity model.5 Nonetheless, I
find his conceptualization useful in constructing the argument to support the pro-
cessualist approach to the growth of scientific knowledge. Thus, I use these three
dimensions: surprisingness, novelty and utility to identify three separate interact-
ing mechanisms which underlie the process of new research routine formation. The
reminder of this section presents detailed examination of these three mechanisms
with regard to microRNAs as a new research routine. Before setting on this task,
however, I embark on a succinct characterization of this newly discovered kind of
RNA particles and overview the main phases that led to establish the corresponding
new research routine.
On the basis of the so called “Crick’s dogma” it has been taken for granted since
1950s [24, p. 1] that DNA is transcribed into RNA, which, in turn, is translated into
protein. At the time preceding the discovery of microRNAs “standard wisdom” [6,
p. 133] was that there existed only three major kinds of RNAs: messenger (mRNA),
transfer (tRNA) and ribosomal (rRNA). The recognition of a new kind of RNA
particles that were not involved in protein translation—so called non-coding RNA—
and acknowledgment of their occurrence in a rich class of organisms was a slow and
long-time process.
MicroRNAs are approximately 20 nucleotides in length that regulate gene expres-
sion by inhibition of translation or mRNA degradation. MicroRNAs may regulate
up to 30% of the human protein-coding genome and control the expression of genes
involved in key biologic processes, including apoptosis, proliferation, differentiation,
and metastasis [2, 4, p. 1]. The microRNAs database—miRBase (v.22.1, October
2018) records 38,589 different microRNAs occurring in 271 kinds of organisms,
including Homo Sapiens with 1917 identified kinds of microRNA.
There are several phases that led to the discovery of microRNAs and their fur-
ther exploration [28]. The relevant line of research has started in 1980 when Vic-
tor Ambros, joined later by Garry Ruvkun, became a post-doc team member of
Robert Horvitz’s team at his laboratory in MIT. Initially, they tried to identify “hete-
rochronic” genes which were responsible for developmental mutations of a nematode
Caenorhabditis elegans (C. elegans), a model organism used in studies of mutations.
By 1989 they were able to identify the functional and later on also the molecular
model. In 1993 Ambros with collaborators reported what is now recognized as the
first microRNA, namely the gen lin-4, which “does not encode a protein”, but rather
“regulates lin-14 translation” [34, p. 843]. A complimentary paper was published by
5 For the details concerning how the values of Simonton’s measures are determined see e.g. [49].
Cognitive Dynamics of Research Routines: Case Study of MicroRNA 139
Ruvkun’s research team [56], who in 2000 demonstrated moreover that the second
microRNA, namely let-7, is highly evolutionarily conserved and—in contrast to lin-4
that occurs only in a small class of organisms—is prevalent in a wide class of species,
including humans [43]. With this result came the recognition that non-coding RNAs
form an important class of RNA particles that regulate expression of other genes.
Right after the standard terminology of “microRNAs” (or “miRNAs”) was coined in
2001 [30, 32, 33] researchers started to explore the role of microRNAs in oncogen-
esis [12]. In 2007 cell-free microRNAs (circulating miRNAs) have been detected in
plasma and serum [52], demonstrating high therapeutic and diagnostic potential of
microRNAs as non-invasive biomarkers.
A synthetic way to illustrate the evolution of microRNA as a research routine is
presented in Fig. 1. In 1993 the giant component in the author collaboration network,
where nodes indicate authors who are linked to their co-authors of their joint pub-
lications, included only 26 authors, who constituted 11% percent of all the authors
who published papers on issues related to microRNAs. In contrast, in 2018 the giant
component included 70% of the authors totaling 27,176 individuals. It not only indi-
cates the widespread diffusion of microRNAs among researchers, but, more impor-
tantly, it demonstrates the fact that microRNA biology has become established as a
sub-discipline of molecular biology of its own.
Prima facie, it appears that the growth of microRNA routine follows a typical
diffusion curve following the inverted S-shape (Fig. 2). In the remainder of this
section I demonstrate that it indeed is a more complex phenomenon by pointing
out three interacting mechanisms related to: breakthrough, novel and utility-oriented
research. As it turns out, the diffusion curve may only be taken to characterize just
one of these mechanisms related to novel research.
While the difference between breakthrough and novel research is well established
and has been epitomized by Thomas Kuhn’s distinction between “revolutionary” ver-
sus “normal” stages for the sake of the research on microRNA as a research routine
Fig. 1 The evolution of microRNA. The two snapshots are based on analysis of MEDLINE data
for microRNA and its MeSH synonyms as MeSH keywords
140 P. Kawalec
I adopt the following distinctions. Breakthrough research creates the new symbolic
representation that later becomes the organizing principle for the expanding social
practices—starting in a single research team and diffusing steadily among other
research laboratories. As demonstrated in my earlier paper [28] microRNA in the
period 1984–2000 has become established as a new symbolic representation, in con-
trast with the earlier predominant view of RNA as coding proteins. Novel research
has expanded this symbolic representation by advancing knowledge of microRNA
biogenesis, complimentary molecular mechanism (RNAi), prediction of target genes,
identification of microRNAs particles and their occurrence in various kinds of organ-
isms, etc. Utility-oriented research applies the knowledge concerning the genesis,
structure, function or types of microRNAs to practical problems of diagnosis and ther-
apy in other areas of research (see also Fig. 6). As mentioned above, utility-oriented
research has commenced as early as 2002 with the publication of [12].
The next section proceeds with a more detailed examination of each of these
mechanisms. This investigation will provide ample evidence against the overly sim-
plistic view that research routine dynamics follows S-shaped diffusion curve. Let
me remark here that the latter view seems to be widely entrenched as it was force-
fully propounded by the founder of scientometric research, namely by de Solla Price
[15–17]. Using an analogy with a growth process that can be observed in multiple
organisms he claimed that the growth of a scientific discipline as evidenced by the
accumulation of the number of publications pertaining to the discipline in question
can be divided into phases each of which instantiates S-shaped curve. In order to
model it, Price used the logit growth model. However, as pointed out recently [5] de
Solla Price’s original idea is more adequately captured by the Gompertz function.6
Now, let me briefly make some remarks in order to explain why I did not decide to
use this function to model the development of microRNA routine. The application of
the Gompertz function, like de Solla Price’s preferred logit growth model, consists in
cutting the relevant dataset into overlapping periods each of which taken separately
6 The interested Reader may consult [5] for details of the Gompertz function.
Cognitive Dynamics of Research Routines: Case Study of MicroRNA 141
Fig. 3 Gompertz curve fitting for microRNA publications. Annual numbers of publications for
microRNA MeSH synonyms in the MEDLINE database. The dates indicate opening and closing
dates for the respective periods. Note that given this partition of the dataset the publication number
for the remaining part, starting in 1997/1998, grows exponentially
The detailed analysis of the three mechanisms underlying the research practice in
molecular biology, as exemplified here with the dynamics of the microRNA research
routine, proceeds in accordance with the three mechanisms described in Sect. 3. These
are: breakthrough, novel and utility-oriented research. I start with a description of
each of these mechanisms separately, and only then comment on issues related to
their interaction.
142 P. Kawalec
For the sake of the present paper I understand breakthrough research as the one
that establishes the symbolic representation of an emergent research routine. In the
case of microRNA this phase of research can be rather precisely identified. It begins
in 1980, when Victor Ambros and Gary Ruvkun join the research team of Robert
Horvitz at MIT and start the research on heterochronic gens of C. elegans.7 The
closing date for this phase is 2000, when Ruvkun’s own research team demonstrated
high conservation of let-7, the second microRNA. Of course, the early research
on heterochronic mutations of C. elegans was continuous with the prior research,
most importantly Horvitz’s contributions in isolating and complete characterization
of cell-lineages of C. elegans mutants and identification of the role of lin-4 in the
process [20].
Undoubtedly, it is cumbersome, if not impossible, to apply the typical sciento-
metric tools to analyze the breakthrough phase of research. None of the two key
publications: Lee et al. [34] and Wightman et al. [56] is captured by typical search
based on a keyword selection. Table 1 provides the complete keyword description
of both publications in MeSH database. Note that none of the keywords on the lists
7 In 1984 both Ambros and Ruvkun moved out to head their own laboratories.
Cognitive Dynamics of Research Routines: Case Study of MicroRNA 143
Fig. 4 Citations of [34]. Citations are grouped in accordance with WoS categories for research
areas, where “basic research” includes: basic biochemistry, cell biology and genetics; “applied
research” includes: applied oncology, medical research experimental, cardiac cardiovascular sys-
tems, hematology; and “commercial research” includes: biotechnology applied microbiology and
pharmacology pharmacy
8 “MeSH” stands for “Medical Subject Headings” and is systematically developed and updated for
indexing articles in PubMed. It identifies 16 synonyms for “microRNA”.
144 P. Kawalec
Fig. 5 Novel research and diffusion curves (2000–2018) for a the number of affiliations, b the
number of publishing authors, and c the number of papers published each year. Analysis of WoS
dataset for “microRNA” as topic
Different explanations are possible. It may turn out that this is an artefact of
how WoS codes the articles. If, on average, each paper is assigned multiple WoS
categories, it may well be that the number of categories will, initially, grow much
faster the number of publications. However, a more plausible explanation is that the
confirmation of the existence and the recognition of high conservation of microRNAs
in 2000 “led to intense research aimed at identifying new members of this family.
This resulted in the discovery of multiple miRNAs across different species of plants
and animals” [9, p. 759–760].
However, a remarkable question remains: How these three mechanisms interact?
One explanation would be that they are largely independent. To some extent this
answer may be supported by the fact that the phase of novel research apparently
kicks-off just right after the breakthrough phase. Yet the same does not hold for the
relation between novel and utility-oriented research as they were advanced in parallel
since 2002.
A more plausible explanation seeks some sort of interaction between the three
phases as obviously developments in each affected the others. A typical linear model
would have it that the outcomes of breakthrough research fed into novel research
and that, in turn, provided input to utility-oriented research. However, this kind of
simple accumulative view is not able to account for distinctly different periods of
rapid growth such as 2008–2012 for novel research (Fig. 2) and 2004–2008 for
utility-oriented research (Fig. 6). Therefore, I would propose here to append the
146 P. Kawalec
Fig. 6 Number of different research areas using microRNA. Analysis of WoS database for
“microRNA” as topic and the WoS categories for research areas
The results of this analysis are presented in Fig. 8. Apparently, some of these
shock impulses, although with varying effects, affect both kinds of research in a
similar way (e.g. in 2005, 2008, 2010, 2013, 2015), while others (such as 2011,
2014, 2016) have adverse effects on each of the two mechanisms. Thus, if one yields
some explanatory power to shock impulses in the dynamics of research routine, then
these results support the view that there are two different mechanisms operating in
the course of the development of microRNA routine.
Fig. 8 Shocks for novel and utility research. Analysis of microRNAs as topic WoS dataset by
means of extension of Fourier series as given by Formula (1)
148 P. Kawalec
5 Concluding Remarks
The arguments presented in the previous sections support the view that research
as a social practice has far greater complexity than assumed by the majority of
scientometric studies. In order to push this point a little further, I proceed to indicate
some pitfalls of the recent proposal to introduce a measure of “paradigm-shifting”
papers. Prabhakaran et al. [45] proposes to use a measure of “the gradient of the
flow vergence” as the sole indicator of the papers, which have transformative impact
in a given research area. This parameter is defined for a network of paper citations,
where indegree and outdegree of a paper i measure respectively the number of its
citations by other papers and the number of other publications cited in the paper i.
Additionally, it takes into account the weight of the citations by considering the so
called eigenvalue of the paper i, which depends on the number of citations it gets
from papers that are themselves highly cited. Thus, flow vergence for a paper i is
defined as the ratio of the difference between incoming and outgoing citations of
a given paper i to the total number of its incoming and outgoing citations and is
appended with its eigenvalue:
indegi − outdegi
W FV i = + eigi . (2)
indegi + outdegi
The gradient of flow vergence from paper j to paper i is, then, defined as a
difference between their respective flow vergences:
△ F V i j = W F Vi − W F V j . (3)
Table 2 Top rank papers in gradient scale. The numbers in parentheses indicate a position of the
paper among 500 top cited papers
Gradient based ranking (500 top Paper source Title
cited ranking)
1 (128) Kota et al. [29] Therapeutic microRNA delivery
suppresses tumorigenesis in a
murine liver cancer model
2 (143) Allen et al. [1] microRNA-directed phasing
during trans-acting siRNA
biogenesis in plants
3 (22) Calin and Croce [13] MicroRNA signatures in human
cancers
4 (5) Lu et al. [38] MicroRNA expression profiles
classify human cancers
5 (144) Lee and Dutta [35] MicroRNAs in cancer
Analysis based on MEDLINE dataset for “microRNA” as MeSH keyword
of the research dynamics and, as a consequence, would call for an elaboration of new
analytic tools that would be more adequate with regard to more complex units of
analysis, such as research routine propounded in this paper. Such advances may be
largely aided by a detailed scrutiny of shock impulses, their sources and interactions
with different mechanisms underlying the dynamics of research routines.
Acknowledgements The author gratefully acknowledges the support of the Polish National
Science Centre (NCN) under the grant no. UMO-2014/15/B/HS1/03770.
References
1. Allen, E., Zhixin, X., Gustafson, A.M., Carrington, J.C.: MicroRNA-directed phasing during
trans-acting siRNA biogenesis in plants. Cell 121(2), 207–221 (2005). https://doi.org/10.1016/
j.cell.2005.04.004
2. Almeida, M.I., Rui, M.R., Calin, G.A.: MicroRNA history: discovery, recent applications,
and next frontiers. Mutat. Res. 717(1–2), 1–8 (2011). https://doi.org/10.1016/j.mrfmmm.2011.
03.009
3. Ambros, V.: A hierarchy of regulatory genes controls a larva-to-adult developmental switch in
C. elegans. Cell 57(1), 49–57 (1989). https://doi.org/10.1016/0092-8674(89)90171-2
4. Ambros, V.: The functions of animal microRNAs. Nature 431(7006), 350–355 (2004). https://
doi.org/10.1038/nature02871
5. Andersen, J.P., Hammarfelt, B.: Price revisited: on the growth of dissertations in eight research
fields. Scientometrics 88(2), 371–383 (2011). https://doi.org/10.1007/s11192-011-0408-8
6. Avise, C.: Conceptual Breakthroughs in Evolutionary Genetics: A Brief History of Shifting
Paradigms. Elsevier, Amsterdam (2014)
7. Bartel, D.P.: MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116(2), 281–
297 (2004). https://doi.org/10.1016/S0092-8674(04)00045-5
8. Bernal, J.D.: The Social Function of Science. Routledge, London (1939)
9. Bhaskaran, M., Mohan, M.: MicroRNAs: history, biogenesis, and their evolving role in ani-
mal development and disease. Vet. Pathol. 51(4), 759–774 (2014). https://doi.org/10.1177/
0300985813502820
10. Binford, L.R.: Archaeology as anthropology. Am. Antiq. 28(2), 217–225 (1962). https://doi.
org/10.2307/278380
11. Buettner-Janusch, John: Boas and Mason: Particularism versus generalization. Am. Anthropol.
59(2), 318–324 (1957)
12. Calin, G.A., Dumitru, C.D., Shimizu, M., Bichi, R., Zupo, S., Noch, E., Aldler, H., et al.:
Frequent deletions and down-regulation of micro-RNA genes miR15 and miR16 at 13q14 in
chronic lymphocytic leukemia. Proc. Natl. Acad. Sci. 99(24), 15524–15529 (2002). https://
doi.org/10.1073/pnas.242606799
13. Calin, G.A., Croce, C.M.: MicroRNA signatures in human cancers. Nat. Rev. Cancer 6(11),
857–866 (2006). https://doi.org/10.1038/nrc1997
14. Campbell, D.T.: Blind variation and selective retentions in creative thought as in other
knowledge processes. Psychol. Rev. 67(6), 380–400 (1960). https://doi.org/10.1037/h0040373
15. de Solla Price, D.J.: Quantitative measures of the development of science. Arch. Int. d’Histoire
Sci. 14, 85–93 (1951)
16. de Solla Price, D.J.: Networks of scientific papers. Science 149(3683), 510–515 (1965). https://
doi.org/10.1126/science.149.3683.510
17. de Solla Price, D.J.: Little Science, Big Science—And Beyond. Columbia University Press,
New York (1986)
Cognitive Dynamics of Research Routines: Case Study of MicroRNA 151
18. Gitschier, J.: In the tradition of science: an interview with Victor Ambros. PLoS Genet. 6(3),
1–4 (2010). https://doi.org/10.1371/journal.pgen.1000853
19. Hacking, I.: The Social Construction of What?. Harvard University Press, Cambridge, MA
(1999)
20. Horvitz, H.R., Sulston, J.E.: Isolation and genetic characterization of cell-lineage mutants of
the nematode caenorhabditis elegans. Genetics 96(2), 435–454 (1980)
21. Hoyningen-Huene, P.: Context of discovery versus context of justification and Thomas Kuhn.
In: Schickore, J., Steinle, F. (eds.) Revisiting discovery and justification: historical and philo-
sophical perspectives on the context distinction, pp. 119–131. Springer, Dordrecht (2006).
https://doi.org/10.1007/1-4020-4251-5_8
22. Ippoliti, E.: Scientific discovery reloaded. Topoi (2017). https://doi.org/10.1007/s11245-017-
9531-3
23. Ippoliti, E.: Building theories the heuristic way. In: Danks, D., Ippoliti, E. (eds.) Building
Theories, pp. 3–20. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-72787-5_1
24. Jarroux, J., Morillon, A., Pinskaya, M.: History, discovery, and classification of lncRNAs. In:
Rao, M.R.S. (ed.) Long Non Coding RNA Biology, pp. 1–46. Springer Singapore, Singapore
(2017). https://doi.org/10.1007/978-981-10-5203-3_1
25. Kawalec, P.: Towards an evolutionary model of science dynamics: generation and production
of scientific knowledge. Zagadnienia Naukoznawstwa 53(4), 405–428 (2017). https://doi.org/
10.24425/118035
26. Kawalec, P.: Philosophical perspectives: the science of science—from inception to maturity.
In: Friedrich, C., Kleeberg, B. (eds.) A New Organon: Science Studies in Interwar Poland,
pp. 521–535. Mohr Siebeck, Tübingen (2020)
27. Kawalec, P.: Integrative methodology: a study of dynamics of scientific knowledge. [in Polish].
Wydawnictwo KUL, Lublin (2018)
28. Kawalec, P.: Transformations in breakthrough research: the emergence of Mirnas as a research
routine in molecular biology. Open Inf. Sci. 2(1), 127–146 (2018). https://doi.org/10.1515/
opis-2018-0010
29. Kota, J., Chivukula, R.R., O’Donnell, K.A., Wentzel, E.A., Montgomery, C.L., Hwang, H.-W.,
Chang, T.-C., et al.: Therapeutic microRNA delivery suppresses tumorigenesis in a murine
liver cancer model. Cell 137(6), 1005–1017 (2009). https://doi.org/10.1016/j.cell.2009.04.021
30. Lagos-Quintana, M., Rauhut, R., Lendeckel, W., Tuschl, T.: Identification of novel genes cod-
ing for small expressed RNAs. Science 294(5543), 853–858 (2001). https://doi.org/10.1126/
science.1064921
31. Langley, P. (ed.): Scientific discovery: computational explorations of the creative processes.
MIT Press, Cambridge, MA (1987)
32. Lau, N.C., Lee, L.P., Weinstein, E.G., Bartel, D.P.: An abundant class of tiny RNAs with
probable regulatory roles in Caenorhabditis elegans. Science 294(5543), 858–862 (2001).
https://doi.org/10.1126/science.1065062
33. Lee, R.C., Ambros, V.: An extensive class of small RNAs in Caenorhabditis elegans. Science
294(5543), 862–864 (2001). https://doi.org/10.1126/science.1065329
34. Lee, R.C., Feinbaum, R.L., Ambros, V.: The C. elegans heterochronic gene lin-4 encodes small
RNAs with antisense complementarity to lin-14. Cell 75(5), 843–854 (1993). https://doi.org/
10.1016/0092-8674(93)90529-y
35. Lee, Y.S., Dutta, A.: MicroRNAs in Cancer. Annu. Rev. Pathol. 4(1), 199–227 (2009). https://
doi.org/10.1146/annurev.pathol.4.110807.092222
36. Liu, Z.C., Ambros, V.: Heterochronic genes control the stage-specific initiation and expression
of the dauer larva developmental program in Caenorhabditis elegans. Genes Dev. 3(128),
2039–2049 (1989). https://doi.org/10.1101/gad.3.12b.2039
37. Losee, J.: Theories of Scientific Progress: An Introduction. Routledge, London (2003)
38. Lu, J., Getz, G., Miska, E.A., Alvarez-Saavedra, E., Lamb, J., Peck, D., Sweet-Cordero, A.,
et al.: MicroRNA expression profiles classify human cancers. Nature 435(7043), 834–838
(2005). https://doi.org/10.1038/nature03702
152 P. Kawalec
39. Nicholson, D.J., Dupré, J. (eds.): Everything Flows: Towards a Processual Philosophy of
Biology. Oxford University Press, Oxford (2018)
40. Nickles, T.: Scientific discovery as a topic for philosophy of science: some personal reflections.
Topoi (2018). https://doi.org/10.1007/s11245-018-9566-0
41. Ossowska, M., Ossowski, S.: Nauka o nauce. Nauka Polska 20, 1–12 (1935)
42. Ossowska, M., Ossowski, S.: The science of science. Minerva 3(1), 72–82 (1964)
43. Pasquinelli, A.E., Reinhart, B.J., Slack, F., Martindale, M.Q., Kuroda, M.I., Maller, B., Hay-
ward, D.C., et al.: Conservation of the sequence and temporal expression of let-7 heterochronic
regulatory RNA. Nature 408(6808), 86–89 (2000)
44. Popper, K.: The Logic of Scientific Discovery, 2nd edn. Routledge, London (2002)
45. Prabhakaran, T., Lathabai, H.H., George, S., Changat, M.: Towards prediction of paradigm
shifts from scientific literature. Scientometrics 117(3), 1611–1644 (2018). https://doi.org/10.
1007/s11192-018-2931-3
46. Schickore, J.: Scientific discovery. In: Zalta, E.N. (ed.) The Stanford Encyclopedia of
Philosophy, Summer 2018 edn. https://plato.stanford.edu/archives/sum2018/entries/scientific-
discovery
47. Simon, H.A.: Models of Discovery and Other Topics in the Methods of Science. Springer,
Dordrecht (1977)
48. Simonton, D.K.: Creativity and discovery as blind variation: Campbell’s (1960) BVSR model
after the half-century mark. Rev. Gen. Psychol. 15(2), 158–174 (2011). https://doi.org/10.1037/
a0022912
49. Simonton, D.K.: Combinatorial creativity and sightedness: Monte Carlo simulations using
three-criterion definitions. Int. J. Creat. Probl. Solving 22(2), 5–18 (2012)
50. Simonton, D.K.: On praising convergent thinking: Creativity as blind variation and selec-
tive retention. Creat. Res. J. 27(3), 262–270 (2015). https://doi.org/10.1080/10400419.2015.
1063877
51. Small, H., Boyack, K.W., Klavans, R.: Identifying emerging topics in science and technology.
Res. Policy 43(8), 1450–1467 (2014). https://doi.org/10.1016/j.respol.2014.02.005
52. Valadi, H., Ekström, K., Bossios, A., Sjöstrand, M., Lee, J.J., Lötvall, J.O.: Exosome-mediated
transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells.
Nat. Cell Biol. 9(6), 654–659 (2007). https://doi.org/10.1038/ncb1596
53. van Fraassen, B.C.: Modeling and measurement: the criterion of empirical grounding. Philos.
Sci. 79(5), 773–784 (2012). https://doi.org/10.1086/667847
54. Volinia, S., Calin, G.A., Liu, C.-G., Ambs, S., Cimmino, A., Petrocca, F., Visone, R., et al.:
A microRNA expression signature of human solid tumors defines cancer gene targets. Proc.
Natl. Acad. Sci. U.S.A. 103(7), 2257–2261 (2006). https://doi.org/10.1073/pnas.0510565103
55. Weisberg, R.W.: Creativity: Understanding Innovation in Problem Solving, Science, Invention,
and the Arts. Wiley, New York (2006)
56. Wightman, B., Ha, I., Ruvkun, G.: Posttranscriptional regulation of the heterochronic gene
lin-14 by lin-4 mediates temporal pattern formation in C. elegans. Cell 75(5), 855–862 (1993).
https://doi.org/10.1016/0092-8674(93)90530-4
57. Znaniecki, F.: Przedmiot i zadania nauki o wiedzy. Nauka Polska 5, 1–78 (1925)
Habit-Based and Goal-Directed Systems:
Knowledge Transfer in Individual
and Social Learning
Robert Lowe
Abstract The present chapter discusses value-based habitual and goal-directed sys-
tems as studied in the animal and human learning literature. It focuses on the means
by which these two systems might interact in knowledge transfer, particularly as it
applies to social learning. Knowledge is conceived here in terms of types of logic
computations as implemented by neural networks. A discussion of dual-process type
structures in the brain is provided as well as neural-dynamic implementations thereof
and considerations for how a perspective of the brain as carrying out logic compu-
tations might be useful for developing the general cognitive capacities of artificial
agents.
1 Background
The relationship between habits and goal-directed systems has been the subject of
much research in the animal (and human) learning community over the last century.
A variety of animal and human learning studies and models thereof have described
the link between habits and cognition (e.g. [1, 2]). Indeed, habits in themselves can
comprise ‘higher’ cognitive phenomena, e.g. habits of thought [3]. Nevertheless,
habits are typically characterized as learned phenomena and “inflexible, slow or
incremental, unconscious, automatic, and insensitive to reinforcer devaluation” [3].
To this end, they have much in common with learning and processing within standard
feedforward neural networks of the type that learn through target (supervised) or
reinforcer feedback. In the animal learning literature, on the other hand, habits are
typically framed as concerning associations between stimuli and responses as they
are reinforced by rewarding or punishing feedback (e.g. [4]).
In this chapter the role of habitual and associative processes are considered with
respect to their role in utilizing or shaping logical functions in the neural networks of
R. Lowe (B)
University of Gothenburg, Forskningsgången 6, Gothenburg 41756, Sweden
e-mail: robert.lowe@gu.se
the brain. Firstly is discussed the relationship between habits, goal-directed systems
and value functions of the type that are formalized in reinforcement learning models
(Sect. 2). Then in Sect. 3 the role of habits (and associative processes) in knowledge
transfer, in relation to boolean logic constructed knowledge, is discussed as it applies
to individual and social learning. In Sect. 4 the relationship between goal-directed
systems and predicate logical function is discussed, again with reference to constitu-
tive associative and habitual processes and how bottom-up and top-down ‘systems’
may facilitate one another. Finally Sect. 5 offers concluding remarks and how appre-
hension of neural network implementation of habit-based and associative structures
might be utilized in artificial, including robotic, agent controllers.
Goal-Directed TD
H a bi t s Habits error
Stimulus-Action-Outcome
Stimulus-Response- Associations Stimulus-Response- Outcome
Associations RPE Associations Value
Expectancy
Outcome Environment
Actions Goals Actions Goals
Habit-Based and Goal-Directed Systems …
Actor
Training Input
signal
Fig. 1 Habits, goals and value functions. Left. Traditional view of habits, connected to a goal-directed system, as value-free (adapted from [9]). Right.
Reinforcement Learning (Actor-Critic) view of habits as temporal difference (TD) value-based—the architecture here has been adapted to include Outcome
Expectancies (potentially modulated by Goal inputs) that, unlike traditional Actor-Critic architectures, directly interface Critic and Actor. The traditional Actor-
Critic architecture, on the other hand, is interfaced only through the reward prediction error (RPE) that forms part of a three-factor learning rule to update both
Critic (Stimulus Value estimates) and Actor (Stimulus-Action associations)
155
156 R. Lowe
Actor-Critic relation and the Stimulus-Action associations that are formed through
a three-factor learning rule that requires reward prediction error (RPE) as compared
to the considered traditional view that entails pure Hebbian stimulus-response asso-
ciations. In Fig. 1, outcome expectancies (Fig. 1, right) may be considered simply as
reward-based goals (as in [14]). If the outcome expectancies are not explicitly repre-
sented, the depiction is comparable to the Actor-Critic architecture of [15, 16]. In this
particular figure, however, the outcome expectations permit representation of differ-
ential outcomes in a manner theorized by Associative Two-Process (ATP) theory—
an animal learning theory that in fact subscribes to three factor stimulus-response
learning (see Fig. 2).
The stimulus-response (S-R) associations of ATP form a value-based habitual
process insofar as they are slow to devaluate relative to goal-directed processes [18].
The significance of the S-R process in the context of ATP theory is the view that the
value-constrained learning of both the S-R process and the outcome/goal-directed
processing interlinks the two types of systems. Habits and goal-directed processes
may not be seen as working separately on this view.
When considered within the context of dual (or multi)-process learning mecha-
nisms such as the Associative Two-Process, value-based habit learning (inflexible,
slow to respond to devalued stimuli) should not be considered a lesser or vestigial
Fig. 2 Associative Two-Process theory and its underpinnings, adapted from [17]. a Standard
behaviourist perspective where S–R associations are learned via the outcome expectation repre-
sentation (value). b Outcome expectation can cue responding following S–E, E–R learning. c Non-
differential outcomes expectations do not provide extra information for cueing responses. d Differ-
ential outcomes expectations provide additional information to stimuli for cueing responses. Here
λ denotes non-differential outcomes; λ1 and λ2 are differential
Habit-Based and Goal-Directed Systems … 157
aspect of learning [19]. Many models of animal learning theorize the existence of
a dual process of memory wherein habit-based learning is embedded. It has been
suggested that the habit system takes control of responding late in learning while the
goal-directed process dominates during early learning that constitutes a period of rel-
ative uncertainty [20]. Such models posit a complementary process variably entailing
expectation, motivational or goal-directed processes (e.g. [4, 21–27]). Habit-based
learning routes may be critical for detailed learning of the relationship between per-
ceptible stimuli or objects and actions [1]. They are also efficiently learned relative to
goal-directed routes that require the learning of multiple associations that link stimuli
to goals (or outcomes) and goals to responses [28]. Furthermore, they may provide
crucial confirmation/facilitation of learned associations that underlie goal-directed
processes [29].
Addressing the question of whether and how habits can guide the learning of
more cognitive-motivational behavior (relevant to goal-directed activity) that has
been posited to develop under dual processes may further clarify the importance
of learning according to habit-based processes. Such cognitive phenomena include
classification-based/category learning [24, 30], affective learning [31–33, 23], flex-
ible switching of behavior [3] and inferential learning [2, 24, 17]. A challenge to
understanding how habitual processing can bring to bear on cognitive processes is
in finding an appropriate level of description of the learning and interactive pro-
cesses that underlie a given dual process system. In the context of an Associative
two-process, habits (Stimulus-Response route) may: (i) compete with, (ii) substitute
for, (iii) facilitate, the action selection influence of goal-directed processing [29] and
vice versa.
fuzzy logic [40]. Deep neural networks (3+ hidden layers) can realize increasingly
complex nested logic functions (fuzzy boolean functions, or Zadeh operators, [41]).
Fuzzy logic thereby provides, for neural network computations, a generalization
of boolean logic to account for the non-linear activations used in deep(er) neural
networks. State function approximators, when implemented by neural networks,
extract critical information from the environment relevant for the agent task. Deep-
Q type neural networks (e.g. [42–44]) typically involve supervised training of the
state function approximator, for example, an autoencoder or convolutional neural
network, whose stimulus/state outputs are then valuated with respect to the viable
actions.
Notwithstanding the early enthusiasm for conceiving of artificial neural networks
as logic computation models of the brain, the more dominant perspective to neural
network function in modern times is to view them as capturing statistical regulari-
ties among related data sets. Deep learning approaches break down ‘big data’ into
features that reliably recur in portions of the data. Multiple layers within the deep
learning neural network permit the capturing of feature compounds whose level of
abstraction scales with increasing receptive field depth. Nevertheless, underlying this
feature extraction view of neural network functionality is the phenomenon of hierar-
chical decomposition of the data set in order to simplify non-linear classification and
regression problems. This is reminiscent of the classical approach to resolving non-
linearly separable (classification) problems, e.g. when constructing neural networks
for resolving linearly non-separable input-output functions (e.g. as for the famous
XOR problem).
Aside from using nested boolean logic function to approximate habit components
(state, actions), dual process reinforcement learning algorithms may be imbued with
predicate logic functions. For example the dual routes (S-R, habit-based; S-E-R,
outcome/goal-based) of the Associative Two-Process (ATP) model, through being
empirically validated by a set of procedures known as transfer-of-control, reveal the
potential for implicit1 transitive inference through which knowledge, in the form of
stimulus inputs and responses classifiable by the outcomes to which they are associ-
ated might be transferred to novel stimulus-response contexts. By way of example,
Fig. 3 schematizes a transfer-of-control scenario. As is typical for the paradigm,
there are three phases. Each of these phases consists of a number of independent tri-
als for learning: presentations of a stimulus, response options, and then a non-negative
‘outcome’ if the correct response is chosen.
The phases break down as follows: Firstly, there is an initial instrumental learning
phase where the two components (S-E and E-R) of the ‘goal-directed’ route can be
learned as well as the S-R (‘habit-enabling’) route. Secondly, a pavlovian (contin-
gency change) learning phase is presented where new S-E associations are made.
Finally, a second instrumental phase is utilized, which uses previously experienced
1 Halford et al. [70] suggests that a transitive inference process that is not based on ‘omnidirection-
ality’, i.e. that ‘outputs’ of an otherwise feedforward process can be used to infer ‘inputs’ is only
implicit and demonstrative of a lower order cognitive process.
Habit-Based and Goal-Directed Systems … 159
Fig. 3 Transfer-of-control paradigm. The conditioning consists of three phases: Phase 1 (Discrim-
ination Training)—an initial instrumental phase where different stimulus-response (S-R) pairings
(S1-R1, S2-R2) yield different outcomes (O1, O2); Phase 2 (Pairing)—a pavlovian learning phase
where new stimuli are presented and associated with previously experienced outcomes; Phase 3
(Transfer Test)—an instrumental transfer phase where the stimuli from phase 2 are re-presented
as are the response options from Phase 1. ATP theory predicts that responding in the transfer test
(phase 3) will be based on already existing S-E and E-R associations learned from the first two
phases where the theorized preferred selections (underlined Rs) are shown in the left diagram and
the S3 → R1 choice process is schematized on the right leading to rewarding outcome 1 (λ1).
Diagram taken from [45]
stimuli and responses but introduces novel stimulus-response pairings.2 This serves
as a test of transfer of the knowledge of the components (S-E and E-R) learned in
the first two phases that provide the relevant building blocks for the S-E-R process
to select the ‘correct’ response in phase 2.
In the specific transfer-of-control example given in Fig. 3, over the first two phases
outcomes (O1 and O2) are common to S1 and S3, and S2 and S4, respectively (given
that in phase 1 the correct responses are made to obtain those outcomes). As a result,
when Phase 3 (transfer test) occurs, since the animal/human has learned S1 and
S3 according to the same outcome (O1)—that is, it has formed S1-E1 and S3-E1
associations—S3 automatically cues the response associated with E1 (learned in
Phase 1), in this case R1 substituting for the external stimulus. No new learning
is required for this in spite of the fact that the subject has not been exposed to the
particular (external) stimulus-response pairing (S3-R1) previously.
ATP postulates, therefore, that by way of a (dual-route) structured learning pro-
cess, a type of transitive inference is possible to find correct responses in the test
phase without the requirement of learning. S3-R1 associations have not been learned
at the beginning of the test phase, but previous experience allows for a transitive
performance of the form A → C (S3-R1) derived from A → B (S3-E1), B → C
(E1-R1).
The above demonstrates a transfer of knowledge by way of predicate logic of the
form succeeds (A, B) imbuing implicit transitive inference as it concerns individual
learning outcomes. In subsequent work [45, 46], we have postulated and empirically
tested the existence of such a form of learning applied to social settings. In [45] we
2 The first and second phase of the transfer of control paradigm can, in fact, be presented in any
order though more standardly the initial instrumental phase is used first.
160 R. Lowe
suggested that viewing the Pavlovian (second) stage as a form of passive observation
in a social context would allow to test hypotheses as to whether participants can
transfer knowledge learned from their own instrumental experience (Phase 1) and
also knowledge learned within the passive social context (Phase 2) to the standard
test phase (Phase 3). Drawing on a review of how value systems are represented
for self and other in joint activities [45], we hypothesized that subjects would be
able to vicariously learn the value observed for the other where the context is non-
competitive. This is to say that participants experience (the presentation of) other’s
stimuli and outcomes as if they were their own and update their value function
accordingly.
Such a ‘social’ transfer of knowledge might be particularly useful during joint
activities or actions [47], particularly if the actions require fast, non-verbal move-
ments [48]. In such situations, knowledge of how the cooperating agent carries out
his/her action may be of secondary utility to information concerning the stimulus
and outcomes that are perceptible to the observing agent. An ATP-based transfer of
knowledge would thereby alleviate a correspondence problem—imitating another’s
actions to achieve an end goal requires knowledge of how one’s own morphology
maps to another. In the above-presented scenario this would not be necessary. Taking
the example in Fig. 3, the observing agent could perceive that S3 → O1, based on
what it can observe, affects the cooperating agent and transfer that knowledge (S3
→ E1) to its own behavior (responses) according to the past relevant associations
made (E1 → R1). The perceiving agent that can then utilize its own transitive bridge
(goal-directed process; S3 → E1 → R1) to correctly infer the S3-R1 (habit-based)
relation that may be utilized by the cooperating agent.
Naturally, the above discussion has centred on imbuing neural networks (as models
of the brain) with logic computations based on feedforward mechanisms. Goal-
directed behavior, however, may entail top-down effects wherein perceptual stimuli
(or states) are not so much passively received but actively searched according to the
particular goals of the time.
In the previous sections, focus has been placed on feedforward neural activity of the
type that imbues associative learning and might be likened to what [49] have referred
to as System 1 processing that are just one part of a dual process (System 1–System 2)
cognitive system in humans. Halford et al. [50, 51] proposed that cognitive capacities
that cover broader dual systems, i.e. System 1-like and System 2-like processes,
can be ranked according to their relational representational complexity and indexed
according to developmental maturation. Seven ranks were posited from ranks 0 to 6
inclusive.
Habit-Based and Goal-Directed Systems … 161
5 Conclusion
References
1. Seger, C.A.: How do the basal ganglia contribute to categorization? Their roles in generalization,
response selection, and learning via feedback. Neurosci. Biobehav. Rev. 32, 265–278 (2008)
2. Phillips, S., Wilson, W.H., Halford, G.S.: What do transitive inference and class inclusion have
in common? Categorical (co) products and cognitive development. PLoS Comput. Biol. 5(12)
(2009)
3. Seger, C.A., Spiering, B.J.: A critical review of habit learning and the basal ganglia. Front.
Syst. Neurosci. 5 (2011)
4. Davidson, R.J., Irwin, W.: The functional neuroanatomy of emotion and affective style. Trends
Cognit. Sci. 3(1), 11–21 (1999)
5. Dickinson, A.: Actions and habits: the development of behavioural autonomy. Philos. Trans.
R. Soc. Lond. B Biol. Sci. 308, 67–78 (1985)
6. Hirsh, J.: The hippocampus and contextual retrieval of information from memory: a theory.
Behav. Biol. 12, 421–444 (1974)
7. Seger, C.A.: The involvement of corticostriatal loops in learning across tasks, species, and
methodologies, In: Groenewegen, H.J., Voorn, P., Berendse, H.W., Mulder, A.B., Cools, A.R.
(eds.) The Basal Ganglia IX, pp. 25–39. Springer, New York (2009)
8. Daw, N.D., O’Doherty, J.P.: Multiple systems for value learning. In: Neuroeconomics: Decision
Making and the Brain, 2nd edn, pp. 393–410. Academic Press, London (2014)
9. Miller, K.J., Shenhav, A., Ludvig, E.A.: Habits without values. Psychol. Rev. (2019)
10. Lowe, R., Ziemke, T.: Exploring the relationship of reward and punishment in reinforcement
learning. In: 2013 IEEE Symposium on Adaptive Dynamic Programming and Reinforcement
Learning (ADPRL), pp. 140–147 (2013)
11. Lee, G., Lowe, R., Ziemke, T.: Modelling early infant walking: testing a generic CPG archi-
tecture on the NAO humanoid. In: 2011 IEEE International Conference on Development and
Learning (ICDL), vol. 2, pp. 1–6 (2011)
Habit-Based and Goal-Directed Systems … 165
12. Li, C., Lowe, R., Ziemke, T.: Crawling posture learning in humanoid robots using a natural-
actor-critic CPG architecture. In: Artificial Life Conference Proceedings, 13, pp. 1182–1190.
MIT Press, Cambridge, MA (2013)
13. Li, C., Lowe, R., Ziemke, T.: A novel approach to locomotion learning: Actor-Critic architecture
using central pattern generators and dynamic motor primitives. Front. Neurorobotics 8, 23
(2014)
14. Rolls, E.T.: On the brain and emotion. Behav. Brain Sci. 23(2), 219–228 (2000)
15. Suri, R.E., Schultz, W.: Learning of sequential movements by neural network model with
dopamine-like reinforcement signal. Exp. Brain Res. 121(3), 350–354 (1998)
16. Suri, R.E., Schultz, W.: Temporal difference model reproduces anticipatory neural activity.
Neural Comput. 13(4), 841–862 (2001)
17. Peterson, G.B., Trapold, M.A.: Effects of altering outcome expectancies on pigeons’ delayed
conditional discrimination performance. Learn. Motiv. 11, 267–288 (1980)
18. Daw, N.D., Niv, Y., Dayan, P.: Uncertainty-based competition between prefrontal and
dorsolateral striatal systems for behavioral control. Nat. Neurosci. 8(12), 1704 (2005)
19. Sun, R.: Interpreting psychological notions: a dual-process computational theory. J. Appl. Res.
Memory Cognit. 4(3), 191–196 (2015)
20. de Wit, S., Dickinson, A.: Associative theories of goal-directed behaviour: a case for animal–
human translational models. Psychol. Res. PRPF 73(4), 463–476 (2009)
21. Mowrer, O.H.: On the dual nature of learning: A reinterpretation of “conditioning” and
“problem-solving”. Harv. Educ. Rev. 17, 102–148 (1947)
22. Trapold, M.A.: Are expectancies based upon different positive reinforcing events discriminably
different? Learn. Motiv. 1, 129–140 (1970)
23. Cardinal, R.N., Parkinson, J.A., Hall, J., Everitt, B.J.: Emotion and motivation: the role of the
amygdala, ventral striatum, and prefrontal cortex. Neurosci. Biobehav. Rev. 26(3), 321–352
(2002)
24. Urcuioli, P.: Behavioral and associative effects of differential outcomes in discriminating
learning. Learn. Behav. 33(1), 1–21 (2005)
25. Lowe, R., Humphries, M., Ziemke, T.: The dual-route hypothesis: evaluating a neurocompu-
tational model of fear conditioning in rats. Connect. Sci. 21(1), 15–37 (2009)
26. Lowe, R., Billing, E.: Affective-associative two-process theory: a neural network investigation
of adaptive behaviour in differential outcomes training. Adapt. Behav. 25(1), 5–23 (2017)
27. Lowe, R., Almér, A., Billing, E., Sandamirskaya, Y., Balkenius, C.: Affective–associative two-
process theory: a neurocomputational account of partial reinforcement extinction effects. Biol.
Cybern. 111(5–6), 365–388 (2017)
28. Braver, T.S., et al.: Mechanisms of motivation–cognition interaction: challenges and opportu-
nities. Cognit. Affect. Behav. Neurosci. 14(2), 443–472 (2014)
29. Urcuioli, P.: Some relationships between outcome expectancies and sample stimuli in pigeons’
delayed matching. Anim. Learn. Behav. 18(3), 302–314 (1990)
30. Urcuioli, P.: Stimulus control and stimulus class formation. In: Madden, G.J., Dube, W.V.,
Hackenberg, T.D., Hanley, G.P., Lattal, K.A. (eds.) APA Handbook of Behavior Analysis, vol.
1, pp. 361–386. American Psychological Association, Washington, DC (2013)
31. Amsel, A.: The role of frustrative nonreward in noncontinuous reward situations. Psychol. Bull.
55, 102–119 (1958)
32. Amsel, A.: Frustration Theory: An Analysis of Dispositional Learning and Memory. Cambridge
University Press, Cambridge (1992)
33. Overmier, J.B., Lawry, J.A.: Pavlovian conditioning and the mediation of behavior. Psychol.
Learn. Motiv. 13, 1–55 (1979)
34. McCulloch, W.S., Pitts, W.: A logical calculus of the ideas immanent in nervous activity. Bull.
Math. Biophys. 5, 115–133 (1943)
35. Rosenblatt, F.: The perceptron: a probabilistic model for information storage and organization
in the brain. Psychol. Rev. 65, 386 (1958)
36. Rumelhart, D.E., Hinton, G.E., McClelland, J.L.: A general framework for parallel distributed
processing. In: Parallel Distributed Processing: Explorations in the Microstructure of Cognition,
vol. 1, pp. 45–76 (1986)
166 R. Lowe
37. Hornik, K., Stinchcombe, M., White, H.: Multilayer feedforward networks are universal
approximators. Neural Netw. 2(5), 359–366 (1989)
38. Navarro-Guerrero, N., Lowe, R.J., Wermter, S.: The effects on adaptive behaviour of negatively
valenced signals in reinforcement learning. In: 2017 Joint IEEE International Conference on
Development and Learning and Epigenetic Robotics (ICDL-EpiRob), pp. 148–155 (2017)
39. Navarro-Guerrero, N., Lowe, R.J., Wermter, S.: Improving robot motor learning with negatively
valenced reinforcement signals. Front. Neurorobotics 11, 10 (2017)
40. Wierman, M.J.: An Introduction to the Mathematics of Uncertainty, 149–150. Creighton
University (2010)
41. Zadeh, L.A., et al.: Fuzzy Sets, Fuzzy Logic. World Scientific Press, Fuzzy Systems (1996)
42. Lange, S., Riedmiller, M., Voigtländer, A.: Autonomous reinforcement learning on raw visual
input data in a real world application. In: The 2012 International Joint Conference on Neural
Networks (IJCNN), pp. 1–8 (2012)
43. Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A.A., Veness, J., Bellemare, M.G., … Petersen,
S.: Human-level control through deep reinforcement learning. Nature 518(7540), 529 (2015)
44. Yang, Z., Xie, Y., Wang, Z.: A theoretical analysis of deep Q-learning. arXiv preprint arXiv:
1901.00137 (2019)
45. Lowe, R., Almer, A., Lindblad, G., Gander, P., Michael, J., Vesper, C.: Minimalist social-
affective value for use in joint action: a neural-computational hypothesis. Front. Comput.
Neurosci. 10 (2016)
46. Lowe, R., Almér, A., Gander, P., Balkenius, C.: Vicarious value learning and inference in
human-human and human-robot interaction. In: 2019 8th International Conference on Affective
Computing and Intelligent Interaction Workshops and Demos (ACIIW), pp. 395–400. IEEE
(2019)
47. Bratman, M.E.: Shared cooperative activity. Philos. Rev. 101, 327–341 (1992)
48. Vesper, C., Butterfill, S., Knoblich, G., Sebanz, N.: A minimal architecture for joint action.
Neural Netw. 23, 998–1003 (2010)
49. Kahneman, D., Tversky, A.: Rational choice and the framing of decisions. J. Bus. 59(4), 251–
278 (1986)
50. Halford, G.S., Wilson, W.H., Phillips, S.: A conceptual complexity metric based on repre-
sentational rank. Psychology Computer Science & Engineering, University of Queensland
University of New South Wales, Queensland 4072, Australia; Sydney, NSW 2052 Australia
(1999)
51. Halford, G.S., Wilson, W.H., Andrews, G., Phillips, S.: Categorizing Cognition: Toward
Conceptual Coherence in the Foundations of Psychology. MIT Press (2014)
52. Halford, G.S., Wilson, W.H., Phillips, S.: Processing capacity defined by relational complexity:
implications for comparative, developmental, and cognitive psychology. Behav. Brain Sci.
21(6), 803–831 (1998)
53. Halford, G.S., Wilson, W.H., Phillips, S.: Relational knowledge: the foundation of higher
cognition. Trends Cognit. Sci. 14(11), 497–505 (2010)
54. Hummel, J.E., Holyoak, K.J.: Distributed representations of structure: a theory of analogical
access and mapping. Psychol. Rev. 104(3), 427 (1997)
55. Hummel, J.E., Holyoak, K.J.: A process model of human transitive inference. In: Spatial
Schemas in Abstract Thought, pp. 279–305 (2001)
56. Doumas, L.A., Morrison, R.G., Richland, L.E.: Individual differences in relational learning and
analogical reasoning: a computational model of longitudinal change. Front. Psychol. 9 (2018)
57. Knowlton, B.J., Morrison, R.G., Hummel, J.E., Holyoak, K.J.: A neurocomputational system
for relational reasoning. Trends Cognit. Sci. 16(7), 373–381 (2012)
58. Hummel, J. E., Holyoak, K. J., Green, C., Doumas, L. A., Devnich, D., Kittur, A., Kalar, D.
J.: A solution to the binding problem for compositional connectionism. In: Levy, S.D., Gayler,
R. (eds.) Compositional Connectionism in Cognitive Science: Papers from the AAAI Fall
Symposium, pp. 31–34 (2004)
59. Lowe, R. Almér, A., Balkenius, C.: Bridging connectionism and relational cognition through
bi-directional affective-associative processing. Open Inf. Sci. 3, 235–260 (2019)
Habit-Based and Goal-Directed Systems … 167
60. Hinton, G.E., Salakhutdinov, R.R.: Reducing the dimensionality of data with neural networks.
Science 313(5786), 504–507 (2006)
61. Frank, M.J., Rudy, J.W., Levy, W.B., O’Reilly, R.C.: When logic fails: implicit transitive
inference in humans. Memory Cognit. 33(4), 742–750 (2005)
62. Amari, S.I.: Dynamics of pattern formation in lateral-inhibition type neural fields. Biol. Cybern.
27(2), 77–87 (1977)
63. Tran, S.N., Garcez, A.S.D.A.: Deep logic networks: inserting and extracting knowledge from
deep belief networks. IEEE Trans. Neural Netw. Learn. Syst. (2016)
64. Rastegari, M., Ordonez, V., Redmon, J., Farhadi, A.: Xnor-net: Imagenet classification using
binary convolutional neural networks. In: European Conference on Computer Vision, pp. 525–
542. Springer International Publishing (2016)
65. Zhou, X., Li, S., Qin, K., Li, K., Tang, F., Hu, S., … & Lin, Z.: Deep adaptive network: an
efficient deep neural network with sparse binary connections. IEEE Trans. Neural Netw. Learn.
Syst. (forthcoming)
66. Goldental, A., Guberman, S., Vardi, R., Kanter, I.: A computational paradigm for dynamic
logic-gates in neuronal activity. Front. Comput. Neurosci. 8 (2014)
67. Vardi, R., Guberman, S., Goldental, A., Kanter, I.: An experimental evidence-based computa-
tional paradigm for new logic-gates in neuronal activity. EPL 103, 66001 (2013)
68. Kiryazov, K., Lowe, R., Becker-Asano, C., Randazzo, M.: The role of arousal in two-resource
problem tasks for humanoid service robots. In: 2013 IEEE ROMAN, pp. 62–69 (2013)
69. Lowe, R., Philippe, P., Montebelli, A., Morse, A., Ziemke, T.: Affective modulation of embodied
dynamics. In: The Role of Emotion in Adaptive Behaviour and Cognitive Robotics, Electronic
Proceedings of SAB Workshop (2008)
70. Halford, G.S., Andrews, G., Wilson, W.H., Phillips, S.: Computational models of relational
processes in cognitive development. Cogn. Dev. 27(4), 481–499 (2012)
Agency of Breath—Beyond Disciplinary
Views on Ritual
Anna Kawalec
Abstract In this article, I focus on a particular ritual that is constitutive for a wide
range of social practices. I examine it with the intention to develop a conception of
ritual that goes beyond traditional patterns and disciplinary boundaries. The proposed
conception of ritual is meant to reveal interdependence between the realm of human
existence and actions, on the one hand, and the realm of the world that is autonomous
with regard to humans, on the other. The conception of ritual as advanced here is
neither subsumed to the dominant categories of Western culture, nor derived from
sophisticated theoretical discussions. For the purpose at hand, I will refer to this
conception as “folk”. However, because the proposed approach has the ambition
of universal validity, I examine selected categories and concepts of philosophy and
anthropology to substantiate my case: (i) the etymological meanings in Sanskrit,
(ii) the Hebrew origins, (iii) the ethnographic facts (on the basis of Alfred Gell’s
research on Muria’s “God playing” and some selected theses of cultural and anthro-
pological theory), (iv) Jerzy Grotowski’s idea of techniques of the performer’s breath
and (v) some points from Karol Wojtyła’s philosophical integrative view of the human
person and agency.
1 Introduction
In this article, I intend to formulate a conception of ritual that goes beyond tradi-
tional patterns and disciplinary boundaries [1]. The conception of ritual as advanced
here is neither subsumed to the dominant categories of Western culture, nor derived
from sophisticated theoretical discussions [12]. For the purpose of this article, I will
refer to it “folk”, although I claim that it has a much wider applicability and is more
comprehensively functional than theoretical speculations determined by the subject
A. Kawalec (B)
Faculty of Philosophy, John Paul II Catholic University of Lublin, Lublin, Poland
e-mail: kawalec@kul.pl
matter and methodology of science framed by the confines of the Western world
[2]. However, because the proposed approach has the ambition of universal valid-
ity, selected categories and concepts of philosophy and anthropology will serve as
material for the course of my investigation (Sect. 2).
One of the most entrenched and, at the same time, schematic Western categories
is the opposition “nature versus culture” [4]. Contemporaneity proliferates with the
trends of cultural thought, forcing the thesis of the absolute universality of human
influence and outputs. Humans interfere with the microcosm and the cosmos and
transform their own bodies, sex and mentality, leaving traces of this interference
as “culture”. Moreover, humans are convinced that they have developed the poten-
tial to transform everything that exists in the surrounding world. This conviction
is further enhanced by innovative technological advances, recurrent scientific dis-
coveries and theories [13] as well as shocking forms of art [29]. Modern Western
people attribute to “culture” the power and governance structure of the entire world.
“Culture” thus fulfils an ideological function, motivating the exceptional activities
of Western humans, who consider themselves the autonomous creators of “culture”,
which is also ingrained in everyday attitudes of Western societies.
The schematic pair of categories “nature-culture” (also in the version “nature is
culture”) delineates the conceptual framework within which the Western concept of
ritual is located. However, I will argue that although it has become (especially owing
to anthropology) the integral component of Western culture, it was not created by
our culture. What matters is that “ritual” is not merely a concept but is the locus
of injunction of a law of nature and the phenomenon that the people of Western
civilization call “culture”, although its sources stem from a realm that transcends
human reality. It is against this background that I attempt to demonstrate that ritual
is the acting law of nature,1 on the assumption that nature means a primordial, prior
and autonomous realm with regard to human acts of transformation.
I advance my argument by elaborating on the claims of anthropologists, especially
Alfred Gell, as well as the concept of agency created by the philosopher Karol
Wojtyła. Gell proposes to focus on the fundamental dimensions of human life (bodily,
mental, social), removing abstract issues as “cultural”. Wojtyła, in contrast, examines
the basic dimension of personal life: the ontological integrity of the person on the
basis of interacting psychosomatic determinations.
The starting point of my discussion of “ritual as the acting law of nature” is the
assumption that the only law of nature is the one that determines the order of the
material and immaterial (the “invisible”—irrespective of the research tools used)
world, and the place of injunction of the material and immaterial dimensions of the
world is constituted by breath of acting beings. My holistic (mainly anthropological)
point of view emphasizes the thesis that the law of nature is a primary and unique
law of the world, in which human being is embedded. The obviousness of this thesis
today is radically undermined by the influence of the Western conceptions of the
nature of human being and the behavioral attitudes of the Western communities [9].
1 The term “acting” is used here in analogy with its meaning in K. Wojtyła’s title phrase “the acting
person”.
Agency of Breath—Beyond Disciplinary Views on Ritual 171
2 I apply this term also to the ethnic groups that share the territory under the influence of the Western
Fig. 1 Allan Houser’s sculpture of Indian (Santa Fe) Photograph Anna Kawalec
In this section I overview selected ideas of ritual. However, their relevance is not
regarded within particular research disciplines, as it apparently would presume one
of the orders of Western culture, but rather they pertain to a new understanding
of ritual advocated here. Thus, I propose to accordingly modify the meaning of
“ritual” adopted in Western studies on culture towards a more universal, “realis-
tic” one. My argument in the reminder of this section focuses on breath as the
fundamental human ritual. I use this paradigmatic example to modify the estab-
lished meaning of “ritual” by referring to (i) the etymological meanings in Sanskrit,
(ii) the Hebrew origins, (iii) ethnographic facts and conceptions (on the basis of
Alfred Gell’s research on Muria’s “God playing” and some selected theses of cul-
tural and anthropological theory), (iv) Jerzy Grotowski’s idea of techniques of the
performer’s breath and (v) some ideas of Karol Wojtyła’s philosophical interpretation
of person and human agency in the context of integration of human beings. Despite
apparent diversity of these sources, they share a common feature, namely they all
indicate the universal character of breath as the fundamental human ritual.
Agency of Breath—Beyond Disciplinary Views on Ritual 173
In the Vedic religion, R.ta (Sanskrit r.tam. , which means order, truth) is the rule of
natural order. R.ta manages the operation of the universe in the micro and macro
dimensions and manages the functions of holistic order: the natural, moral and sac-
rificial. In the hymns of the Vedas, R.ta, as “one of the most important religious
conceptions of the Rig Veda”, indicates that “from the point of view of the history of
religious ideas we may, in fact we must, begin the history of Hindu religion at least
with the history of this conception” ([3], pp. 12–13). R.ta is derived from the Sanskrit
verb root r., which means “to go, move, rise, tend upwards”, and the derivative noun
r.tam is defined as “fixed or settled order, rule, divine law or truth”. R.ta is translated
literally as “that which has moved in a fitting manner”, abstractly as “universal law”
or “cosmic order”, or simply as “truth”.3 In the context of Vedic religion, the features
of nature that either remain constant or occur on a regular basis were seen to be a
manifestation of the power of R.ta in the physical cosmos [18]. In the human sphere,
R.ta was understood to manifest itself as the imperative force behind both the moral
order of society and the correct performance of Vedic rituals.4 The Sanskrit source
points to the fundamental role of the natural order of universe, while human laws
supervene on that order.
The sources of Hebrew orthopraxis, which also influenced Christianity, include some
interesting practical indications5 :
1. Take a breath and breathe it out.
2. Do it again, slowly, and try to mean it.
3. Breathing—of all things, maybe we take it most for granted.
The question, “Do we ever wonder why we are built this way, this soft machine
of ours always pumping oxygen in and out? In sadness, we breathe heavy sighs. In
joy, our lungs feel almost like they will burst. In fear, we hold our breath and have to
be told to breathe slowly to help us calm down. When we’re about to do something
˙
3 “AŠA, “truth” in Avestan, from Indo-Iranian *r.tá-, a neuter noun having the same meaning. The
word is attested in Old Persian as a.rta and in Old Indian as r.tá-”. See ([24], pp. 168–182): (1) *r.tám
man- “to think (of) truth” (Y. 31.19 and Rigveda); (2) *r.tám vaźh- “to drive the truth” (Y. 46.4
and Rigveda): “The religious poet drives the true words of his hymn as the charioteer the horses”.
For more information see http://www.iranicaonline.org/articles/asa-means-truth-in-avestan. (Last
accessed 17 November 2017).
4 Ibidem.
5 https://rabbitroom.com/2011/08/is-the-name-of-god-the-sound-of-our-breathing/. (Last accessed
10 April 2017).
174 A. Kawalec
hard, we take a deep breath to find our courage. When I think about it, breathing
looks almost like a kind of praying”.6
Hebrews ask, “What if the name of God is the sound of breathing? ‘In-out’”.
They conclude, “How generous of God to choose to give himself a name that we
can’t help but speak every moment we’re alive. All of us, always, everywhere, waking,
sleeping, with the name of God on our lips. When a baby is born—newly arrived
on planet earth, must they take their first breath, or rather speak the name of God
if they are to be alive here? On our deathbed, do we breathe our last breath? Or is
it that we cease to be alive when the name of God is no longer on our lips?” For
them, each negation of the existence of God affirms his existence. God is everywhere
and always, in the universe and inside me: “Breathe in. Breathe out. ‘He does our
praying in and for us, making prayer out of our wordless sighs… the word that saves
is right here, as near as the tongue in your mouth’…” (Romans 8:28, 10:8, The
Message). These assumptions are not confined to the Hebrew religious traditions—
they form the foundation of religious experiences of other religious communities,
such as Christians and the aforementioned Indian. They express the constitutive law of
their existence, nonseparable from the laws of nature as materially and symbolically
manifested in the ritual breath.
Richard Schechner propounds the cultural and performative meaning of ritual draw-
ing upon the accomplishments of Victor Turner [27] and Milton Singer [25]. As a
result, Schechner depicts the evolution of ritual and its genre as a tree [22]. Branches
and twigs mean ritualization among people: social ritual (everyday life, sport, pol-
itics), religious ritual (rites, liturgy), esthetic ritual (codified forms, ad hoc forms);
branches—social ritual (primates without humans); boughs—genetic and free behav-
ior (birds, mammals); trunk—genetically determined behavior (fish, insects), and
roots—ritualization.
Schechner thus localizes ritual between evolutional origins and social activities.
He concludes that every activity of the elements of the symbolic tree of nature of
ritualized behaviour expresses the law of ritualization. However, we do not know
the genre of this law. For Schechner at that time of writing The Future of Ritual it
may have been a social rule. For the later Schechner, however, it may have been a
performative rule,7 according to his words: “They [rituals] are ambivalent symbolic
actions pointing at the real transactions even as they help people avoid too direct
a confrontation with these events. Thus rituals are also bridges—reliable doings
6 Ibidem.
7 See [23]. “Being” is existence itself. “Doing” is the activity of all that exists, from quarks to
sentient beings to supergalactic strings. “Showing doing” is performing: pointing to, underlining,
and displaying doing. Explaining ‘showing doing’“ is performance studies” (p. 28).
Agency of Breath—Beyond Disciplinary Views on Ritual 175
8 ([22],
p. 223). For more on the subject as the agent of the performative act, see [15].
9 Turner’s idea is further elaborated in [5]. Interestingly, the paper has the same title as Turner’s but
does not list it among the references.
10 Turner follows here [10].
176 A. Kawalec
The Western conceptions of ritual, as well as of nature, tend to emphasize the domi-
nant role of human beings, and recognize ritual itself as a mere expression of a human
or social order. Marylin Strathern perceives British sociality as the constructive fact
in her monograph After Nature: English Kinship in the Late Twentieth Century [26].
While she initially formulates a claim about society as the construction of the world
and person, the main thesis of the book is conveyed by the formula, “The nature is
the construction of the human being” ([26], pp. 192–193). According to Strathern,
there is currently a person-social construct and the family as the social construct;
these are the subject of anthropology and social science.
Strathern seems to have committed the error suggested by Gell for Durkheim [7]:
she redefines social constructions, in this case leading to culture. Moreover, it seems
that she reverses the order of social relations (for which the ontological basis is real
persons) and their products or their institutional forms [12].
Perhaps the nearest approximation to the idea of ritual as the acting law of nature,
propounded in this paper, can be found in the publications of a British anthropologist
Alfred Gell concerning “swinging gods”. He observed that at least some of the
traditions we are already acquainted with contain the motifs of dancing gods or
swinging and rocking deities (Fig. 2). Moreover, this theme is common to all Hindu
traditions since at least 900 A.D. and its entrenchment is manifested by cultivation
of a festival at which the gods are ritually “swung”. This festival lasts approximately
10 days, held in May/June, and takes place in the middle of territory of India, so
called “pen karsana” ([6], pp. 219–248).
Among the modes of “play” there is “hook swinging”, comparable to other
forms of self-mortification (flagellation with chains, piercing of the body, picking
up the eyes, fire-walking). All swinging relates to the experience of vertiginous play
(between bodily equilibrium and disequilibrium and states of consciousness). In
all swinging, there is a common element of self-surrender to a loss of individual
equilibrium. It seems that human beings thereby imitate the behavior of their tradi-
tional gods. During swinging, gentle rocking induces peaceful repose, and energetic
rocking motions induce exhilaration, ecstasy and compulsion. Thus, the swing is an
artefact whose use is mainly for the modification of mental states.
Gell advanced an original explication of the tradition of swinging. There are three
interpretations of allied states (trans states) of possessed persons: (1) generalized
“sensory overload”, (2) chemical changes in the body as a result of over-breathing
and (3) the result of rhythmic stimulation by sound or light pulses at a frequency
approximating that of natural alpha brain rhythms [6]. According to Gell, there are
Agency of Breath—Beyond Disciplinary Views on Ritual 177
Fig. 2 Panel a Shiva dancing. India. Photograph G. Solecki, A. Pi˛etak. Courtesy of the Directorate
of the National Museum in Szczecin, Poland; panel; b Medium on the swing. Photograph Alfred
Gell (The Gods at Play). Courtesy of Simeran Gell
two basic elements of trance induction: first, the modification of normal posture and
muscle tone (attention while standing rigidly, extension of the neck, stiffening of
otherwise flexible fingers and arms, staring and yawning); second, voluntary and
later uncontrollable shaking and trembling, shaking of the head, rapid movements
of the forearms up and down, hand-flapping, dancing, and twirling in the air.
The British anthropologist participated in a Hindu festival and noted that the Muria
were possessed without specific sounds and light, without visible modifications of
breathing, and without special stimulus. Therefore, Gell proposed the interpretation
of possession as the result of material incentives (as above) but also as immaterial
incentives: as the intention of a person to be possessed.11 According to Gell, the
Muria trance is much more complex than this material possession in that it is the
body itself. He saw the trance becoming a vehicle (like a journey by horse, a swing,
or a ride on an elephant). Through this trance, the subject can rediscover (across the
trance-gap) between intention and experience. Gell explained that the Body is the
horse, and the Soul is the rider. We can ask, what is the world in this metaphoric
structure?
11 The category of intention is central in Gell’s conception of nexus. See [8, 16]. In the context of
Gell’s theory, it seems that the intention is the intermediary material and immaterial component of
the (human and environmental) world.
178 A. Kawalec
12 Gell underlined this “equilibrium sense” as consisting of the gravity- and movement-sensing
organs in the inner ear (semicircular canals in the vestibule), which are connected to a number of
sites in the central nervous system, which together constitute the vestibular system. The vestibular
system monitors equilibrium and controls the muscles of the head, neck and upper body as well as
the posture.
Agency of Breath—Beyond Disciplinary Views on Ritual 179
In women, correct respiration has a definite abdominal phase, although the upper
thoracic element is slightly different than in men. The actor should practice different
types of respiration in various positions and physical actions (such as acrobatics). It
is necessary to accustom oneself to total respiration. This helps a person to control
herself and to be ready to control the functioning of the respiratory organs.
The most common actors’ practice is breathing. Grotowski very much appreci-
ated the Hindu tradition and its daily practices and rituals.13 It is well known that
the different schools of yoga—including Hatha Yoga—demand the daily practice
of respiratory techniques to control and exploit the biological function of breath-
ing. These practices have become automatic. Grotowski claimed that an actor needs
many exercises for the respiratory process and proposed some ways to verify total
respiration. First, respiration in the horizontal position (Fig. 3).
Next, from classical Chinese theatre, when standing, place the hands on the two
lowest ribs. Inspiration must give an impression of beginning in the very spot where
the hands are placed and, continuing through the thorax, produce a sensation that
the air column reaches right up to the head. This means that when breathing in, the
abdomen and lower ribs dilate first, followed in smooth succession by the chest.
During (correct) total respiration, “Do not store up or compress too much air!” ([19],
p. 150).
In his work with actors, Grotowski used various techniques of breathing so that
each individual discovered the proper place of crossing of the biological and the
psycho-cultural determinations. It seems that for Grotowski—“the teacher for Per-
former”—breath was a bridge over the hiatus of the external and internal human
environments, the physiology, the social dimension of the artist and her spirituality.
It was in breath that he attempted to discover the law of nature exemplified in the
individual.
Arguably, among the philosophical conceptions of ritual as the acting law of nature
the one that seems most pertinent here is the most integral philosophical conception
of the human person advanced by Karol Wojtyła.14 In his research, he used mainly
phenomenological methods, which means that he took detailed description of sensory
experiences and objects as the starting point for his analyses. He was essentially
interested in human activities. Wojtyła understood human person as the only acting
being. He proposed an idea of a human being as a person, a being who is integrating
physical, sensual and spiritual dimensions, with the latter (embracing volitional and
conscious character) being ontologically prime.
Wojtyła proposed systematic research on the person as the way from her expres-
sions to the inner realm, i.e., to the source of these expressions. According to Woj-
tyła, there is no human act without its agent/subject, and vice versa: there is no agent
without her outer expression. Wojtyła connects the ontological basis of the human
as person with the understanding of the relation: person-acting. He wrote of “the
most deeply hidden roots of the dynamism of man, and more particularly the roots of
efficacy that is the key for the understanding” of the person ([31], p. 76). According
to Wojtyła, nature defines the subjective basis of acting. However, this “nature” has
two meanings: nature as biological life and nature as the essence of a specific indi-
vidual being. In Wojtyła’s philosophical context, Aristotle’s hylemorphism seems to
be transcended by the concept of the spiritual essence of the human being. Wojtyła,
as a Western philosopher, underlined the specificity of human existence among other
kinds of being: rationality and self-consciousness, voluntarity and the resulting skills
of the subject, self-governance, self-possession, and self-determination.
Wojtyła indicates that the acting of the integral human being (vegetative-sensual,
covetous, voluminous and rational) is the realm of the crossing of the world of nature
as the material and immaterial world, while the individual act of the human person
witnesses this realm. Thus, the law of the existing world is—according to Wojtyła—
a harmonious interaction between the acting human person (as a free and rational
agent) and the acting nature (as an objectively existing reality).
Etymologically, the term “nature” is derived from the Latin verb “nascor” (to be
born). It denotes literally everything that is going to be born or is contained in the
fact itself of birth as its possible consequence. This literal meaning is basic in Gell’s
approach to the difference of nature-culture. For him, the notion of culture is the high
abstractive notion specific to the Western tradition. “Culture” includes everything that
we may imagine, and what we want. Therefore, the person is also the culture.
The anthropological research approach of Alfred Gell indicates that a tribe’s
thinking and understanding of the world, the self, and the environment has an integral
character (for example, totemic thinking). Therefore, the soul and body, ancestors,
and elements of the environment are integrated. This is the individual person [8].
14 Since1978, he has been known as the Pope John Paul II. Wojtyła’s main philosophical work is
The Acting Person [31].
Agency of Breath—Beyond Disciplinary Views on Ritual 181
According to Wojtyła, this is nature, which determines the ways of acting (vegeta-
tive: the plants; vegetative and sensual: the animal; vegetative, sensual, rational, and
volitional: the person). Rational and volitional dimensions are determined by others;
mutually, the vegetative and sensual determine the higher dimensions (however, the
person can self-determined only by the volitional-rational dimension of the subject).
According to Wojtyła, all dimensions—physical, psychic, and spiritual—create the
human acting.
According to Wojtyła and Gell, the person is the agent and performer in the
same process of action, which is the dynamic action-motion synthesis. For Wojtyła,
various forms of the disintegration of the person may be defined as purely somatic.
Such disintegration may take the form of the absence of a certain organ or member
of the human body, resulting in the inability to make certain motions, or it may be
due to an absence of some purely somatic reactions or skills leading to impediments
or disorders of human action. For example, “A man who has lost an arm or breath…
is subject to very definite limitations or difficulties in his acting […] BUT these
are external in the sense that they do not in themselves distort his consciousness or
prevent his self-determination. On the contrary, very often a human being with a high
degree of somatic disintegration may represent a personality of great value” ([31],
p. 215). Thus, according to Wojtyła, the person is the integration of the somatic,
psychic and spiritual powers, but at the same time, the person is self-determined
(only by subjective power), whereas the breath of humans is the experienced and
unknown commonplace external and internal world, the injunction of the somatic,
psychic, and spiritual power of agency.
3 Concluding Remarks
If we suspend the superstition concerning the distinction between culture and nature,
ritual will lose its symbolic character (imposed upon it by the Western people). Ritual
will become an expression of nature (a repetitive repetition of the breath), determining
the actions of the breathable creatures. Ritual (in the sense of R.ta derived from the
Sanskrit), in this sense, is therefore the commonplace of the external reality for
breathable beings. Ritual as R.ta justifies the natural and inviolable character of the
whole reality irrespective of the labels given to it by various civilizations, religions,
and disciplines.
The presented examples of various ideas aimed to support this conclusion, but
they may be less (Gell, Rappaport, Wojtyła) or more (Schechner, Strathern) limited
by the Western distinction of nature-culture. As argued in the previous section, it is
Sanskrit that instantiates the idea of the most realistic conception of ritual. The pre-
sented examples of various ideas come from various fields of life of human being and
social. But all of them are integrated by the above mentioned theses and pictures that
stem from different areas of human and social life. Their all share a non-scientistic
perspective, and foremost, before taking the fact of human individual and social exis-
tence as correlated with (intangible) law of nature. This correlation can be perceived
182 A. Kawalec
Acknowledgements The author gratefully acknowledges support from the grant NCN
2013/11/B/HS1/03961, awarded by the National Science Centre in Poland.
References
1. Agier, M.: Epistemological decentring: at the root of a contemporary and situational anthro-
pology. Anthropol. Theory 16(1), 22–47 (2016). https://doi.org/10.1177/1463499616629270
2. Apter, A.: Ethnographic X-files and Holbraad’s double-bind: reflections on an ontological turn
of events. HAU: J. Ethnogr. Theory 7(1), 287–302 (2017). https://doi.org/10.14318/hau7.1.021
3. Bloomfield, M.: The Religion of the Veda: the Ancient Religion of India: from Rig-Veda to
Upanishads. Adamant Media Corporation, Boston (2005)
4. Descola, P.: Beyond Nature and Culture. The University of Chicago Press, Chicago (2013)
5. Geertz, A.W.: Brain, body and culture: a biocultural theory of religion. Method & Theory in
the Study Relig. 22(4), 304–321 (2010). https://doi.org/10.1163/157006810X531094
6. Gell, A.: The Gods at play: vertigo and possession in Muria religion. Man 15(2), 219–248
(1980)
7. Gell, A.: The Anthropology of Time: Cultural Constructions of Temporal Maps and Images.
Berg Publishers, Oxford (1992)
8. Gell, A.: Art and Agency: an Anthropological Theory. Oxford University Press, Oxford (1998)
9. Goodman, A.H., Heath, D., Lindee, M.S. (eds.): Genetic Nature/Culture: Anthropology and
Science Beyond the Two-Culture Divide. University of California Press, Berkeley (2003)
10. Grimes, R.L.: Beginnings in Ritual Studies. University Press of America, Washington, DC
(1982)
11. Grotowski, J.: Towards Poor Theatre. Simon & Schuster, New York (1968)
12. Holbraad, M., Pedersen, M.A.: The Ontological Turn: an Anthropological Exposition.
University Press Cambridge, Cambridge (2017)
13. Hornborg, A.: The political economy of technofetishism: agency, amazonian ontologies, and
global magic. HAU: J. Ethnogr. Theory 5(1), 35–57 (2015). https://doi.org/10.14318/hau5.
1.003
14. Huxley, J.: Introduction. Philos. Trans. R. Soc. Lond. Ser. B, Biol. Sci. 251(772), 249–271
(1966)
15. Kawalec, A.: Comic agents: from a poetic to an anthropological paradigm of comedy. Agathos:
Int. Rev. Humanities & Soc. Sci. 7(1), 7–23 (2016a)
16. Kawalec, A.: Person and Nexus: Alfred Gell’s anthropological Theory of Art (in Polish).
Wydawnictwo KUL, Lublin (2016b)
17. O’Donnell, J.K. (ed.): Here, Now, and Always: Voices of the First Peoples of the Southwest.
Museum of New Mexico Press, Santa Fe, NM (2001)
18. Oldenberg, H.: Die Religion des Veda. W. Hertz, Berlin (1894)
19. Osiński, J.: Jerzy Grotowski. Źródła, inspiracje, konteksty. Słowo/obraz terytoria, Gdańsk
(2010)
20. Owings, A.: Indian Voices: Listening to Native Americans. Rutgers University Press, New
Brunswick, NJ (2011)
21. Rappaport, R.A.: Ritual and Religion in the Making of Humanity. Cambridge University Press,
Cambridge (1999)
22. Schechner, R.: The Future of Ritual: Writings on Culture and Performance. Routledge, London
(1995)
23. Schechner, R.: Performative Studies: an Introduction. Routledge, London (2013)
Agency of Breath—Beyond Disciplinary Views on Ritual 183
24. Schlerath, B.: Awesta-Wörterbuch. Vorarbeiten II: Konkordanz. Otto Harrassowitz: Wiesbaden
(1968)
25. Singer, M. (ed.): Traditional India: Structure and Change. The American Folklore Society,
Philadelphia (1959)
26. Strathern, M.: After Nature: English Kinship in the Late Twentieth Century. Cambridge
University Press, Cambridge (1992)
27. Turner, V.: The Ritual Process: Structure and Anti-Structure. Routledge, London (1969)
28. Turner, V.: Body, Brain, and Culture. Zygon 18(3), 221–245 (1983). https://doi.org/10.1111/j.
1467-9744.1983.tb00512.x
29. Vitz, R.: Reforming the Art of Living Nature, Virtue, and Religion in Descartes’s Epistemology.
Springer, Cham (2015)
30. Wall, D., Masayesva, V.: People of the corn: teachings in hopi traditional agriculture, spirituality,
and sustainability. Am. Indian Q. 28(3), 435–453 (2005). https://doi.org/10.1353/aiq.2004.
0109
31. Wojtyła, K.: The Acting Person Trans. A. Potocki. Reidel, Dordrecht (1979)
From Habits to We-Intentionality:
Rituals as Social Habits
Raffaela Giovagnoli
Abstract Habits and rituals represent a relevant field to investigate human behavior
also in comparison with non human animals and nature in general. We move from
very interesting perspectives that cross philosophy and neuroscience to establish a
continuity between individual and social habits and to clearly isolate what makes
human rituals peculiar in the field of social ontology. Therefore, we introduce a
general and brief discussion on rituals in different disciplines before to focus on
the actuality of the Aristotelian notion of habit in relationship with some studies
in neuroscience that can be extended to the social context. The perspective of We-
Intentionality introduced by John Searle embeds the dimensions of habits that ground
cooperation and is very useful to describe the peculiarity of human rituals.
1 Introduction
R. Giovagnoli (B)
Faculty of Philosophy, Pontifical Lateran University, Vatican City, Italy
e-mail: giovagnoli@pul.it
Habermas conception of rituals moves from the analysis of the normativity character-
izing the background of knowledge and action. Life-world is an important key-notion
characterizing Habermas post-metaphysical thought, namely the choice for the onto-
logical primacy of ordinary language instead of consciousness or any principle based
on the so-called first-philosophy [4]. It represents the horizon of human perceptive
and cultural experience, namely the background from which we raise our validity
claims and exchange our reasons for knowing and acting. Life-world reveals itself in
the dialogical situations where the pragmatic structure of communication acquires its
fundamental role. Language must be pragmatically interpreted, namely starting from
the communicative praxis of the subjects who, while trying to dialogically solve their
questions, are also able to learn from their experiences. Communication possesses
a constitutive role: it continuously forces the participants to undertake a positions
for or against available validity claims. So, they find themselves always running the
risk to be contradicted and so, they have the possibility to correct their opinions in a
autonomous and rational manner [5].
As Husserl himself has shown in his latest thought, life-world has a double nature,
namely it is both the epistemological and ontological background for science and
human activities in general. Habermas proposes to intend lifeworld as structured by
linguistic practices, so the problem is how we can intend these practices that open
and project a world as something that happens in the world. It si a fact that we can
producing advances in our daily life only by facing the contingency of the events.
Consequently, there exist a relation between our practices and these very events,
From Habits to We-Intentionality: Rituals as Social Habits 187
which we can easily grasp in case of failures of our projects. Beyond the traditional
primacy of the phenomenological consciousness or a religious interpretation of the
world, Habermas proposes a third solution: the transcendental spontaneity of the
Kantian Ego manifests itself in the very practices where the reproduction of life-
world is connected from the inside with our mundane processes of learning. These
learning processes take place by virtue of an interactive exchange among intentional
references to the world, reciprocal assumption of perspectives, use of a proposition-
ally differentiated language, instrumental reasoning and cooperation. We find here a
close connection between human agency and world through communication, which
represents the epistemological dimension characterizing our learning and cognitive
development. Nevertheless, It is important to stress that Habermas does not suggest
a merely “cultural” solution because he recognizes the surprising results of the bio-
sciences, psychology and cultural sciences that could play a central, heuristic role in
the development of the natural history of human life.
In his latest thought Habermas maintains that human agency acquires normativity
not only from ordinary communication but also from extra-ordinary communica-
tion. Life-world can be also described as a “space of reasons” (to use the metaphor
of Sellars) which can acquire a social structure as Robert Brandom has shown in his
important work [6]. Habermas aims at describing the intersubjectivity that is typical
of human beings who are able to have collective intentionality in the form of cooper-
ation. Humans have their unique way to interact so, even though non-human animals
have their rituals, humans have also their peculiar way to constitute and to do their
parts in ritual practices.1 According to Tomasello: “The understanding of persons
as intentional agents at one to two years of age—in ways that show sharpness and
perspective—thus inaugurates the development of uniquely human skills of social
cognition” [7]. This understanding involves appreciation of the “original normativ-
ity” of actions because an intentional agent’s action—either self or other—may be
judged as successful or unsuccessful. One-year-olds have thus entered, at least in
a nascent and implicit way, the space of reasons involving normative judgements.
But it is even more relevant for our discussion that children of this age also come to
appreciate that shared intentionality and collective practices create “derived norma-
tivity”. This is a more deeply social sense of normativity, which entails the use of
symbols, artifacts, and other culturally constituted entities. These entities are invested
with normativity through the actions of intentional agents and their attitudes: this is
the way “we” use a symbol or tool. This is the way it “should” be used; this is its
“function” for us, its users.
Derived normativity is most readily apparent in the use of linguistic symbols and
material artifacts such as tools and toys. As Lorenzo Magnani shows, the role of
habits in living organisms—animals and humans alike. Gordana Dodig Crnkovic wonders what the
essential difference between animals and plants results in animals (with humans) having ritual,
patterned, repeating, symbolic, indirectly purposeful communication procedures while we cannot
detect rituals in plants—even though plants exhibit rich forms of communication. Can it be that
animals (including humans) move, and that movement as a part of communication plays an important
role in rituals?
188 R. Giovagnoli
habits, rituals and symbols in human cognition can be elucidated by the use of two
concepts: “manipulative abduction” and “disembodied distributed mind”. In this
sense, human mind can delegate to external representations suitably stored in the
environment important cognitive functions. These representations:
(…) can be merely mimetic (when they represent something already internally present in the
mind, for example a triangle depicted in a dashboard), but also poetic, when external repre-
sentational aspects are manipulated to the aim of discovering something new by exploiting
the specific constrains of the materiality of the external “mediators” chosen. [8]
As Durkheim has shown, rituals reveal their function in the re-generation of soli-
darity as well as in the self-thematization of the communitarian identity. Rituals are
self-reflexive practices that deserve to stabilize the compactness of the social group.
From Habits to We-Intentionality: Rituals as Social Habits 189
In order to grasp the relationship between habits and rituals, we need to present
a plausible sense for the notion of habit, which goes beyond the mere repetitive
behavior or routine. Starting from Latin, there are two meanings for the English word
“habit”. The first is Habitus, that entails a deliberate disposition to act; the second
is Consuetudo, that implies the constant repetition of an event or behavior without
deliberation. The traditional philosophical sense of habit (Habitus) is introduced by
Aristotle to characterize the notion of “virtue”. Virtue can be considered as a habit in
the sense of a disposition to deal with good or bad emotions and tendencies. Aquinas
190 R. Giovagnoli
moves from the Aristotelian view and maintains that habit is not a mere potency (i.e.
a capacity) because it makes us able or unable to act in a good or wrong way. Dewey
maintains that it is a human activity influenced by previous activity: we can say that
it is acquired in the process of socialization and entails a certain order and a certain
system of minor elements of action. Habits can be seen as dynamic dispositions that
work in a subordinate form when they are not the dominant activity at a certain time.
The meaning of the term habit (Consuetudo), as constant repetition of an event or
a behavior, refers to a mechanism that can be described in the fields of natural and
social sciences. Standardly, we assume that this mechanism mostly develops from
the repetition of acts and behaviors and, in the case of human events, it is acquired
through training. Aristotle, conceived this notion of habit as a mechanism that is
analogous to natural mechanisms, and somehow guarantees the uniform repetition
of facts, acts, or behavior by eliminating or reducing effort and fatigue and so by
making them pleasant. Habit as repetition without reasoning is also exemplarily
described by Pascal and Hume. Bergson presented a relationship between habits and
moral obligations: moral obligations can be interpreted as social habits to favor social
life and social order. Metaphysical interpretations of the notion of habit are offered
by Main de Biran, Hegel and Ravaisson. In this case, we can observe that habit
has a strong connection with religious views. According to Hegel, habit is the most
essential thing relevant to the existence of spirituality of the individual subject; the
subject can exist as a concrete subject and as ideality of the soul, namely the religious
content can belong to himself (with his own soul). The metaphysical perspective of
Ravaisson presents a notion of habit as a law of grace that has the important result
to consider nature as spirit and spiritual activity, because it demonstrates that spirit
can turn into nature and vice versa.
A habit is not only a mere automatism or a repetitive behavior, but also a stable
disposition for action (practical skill), that implies the relationship between automa-
tism and flexibility. We can recall the famous aristotelian difference between techne
and praxis, which entails important consequences for collective agency [11]. The
difference between habits and automatism or simple routines is that the former give
control over actions, while the latter don’t [12]. According to this view, that crosses
philosophy and neurobiology, the habit is a “stable disposition for self-development”.
From Habits to We-Intentionality: Rituals as Social Habits 191
We can observe habits according to an individual mode or I-mode because they are
idiosyncratic as regards personal behavior. Habits have a very important function in
individual life because they reduce the complexity of daily life; they make our daily
life easier and pleasant. Naturally, we can control habits concerning the satisfaction of
our basic natural needs. Depending from natural and social environment, we develop
different habits which organize the way to satisfy our human needs.
For example the singer, who can easily reproduce notes (a mostly theoretical
habit) and whose voice appropriately responds to the reading (a technical habit), is
able to give her own interpretation. In the case of behavioral and technical habits,
they imply the availability of motor skills for complex activities, as well as the
modulation of tendencies and desires to positively respond to conscious and rational
goals. Therefore, they entail the performance of habits-as-routines, but their critical
characteristics go beyond their motor aspects.
Habits-as-routines and habits-as-learning have a different relation to conscious-
ness. Habits-as-routines represent a fully unconscious performance. Habits-as-
learning reduce or eliminate consciousness of basic elements of the action in order to
concentrate on higher goals, while preserving at all times the possibility of recovering
them for conscious attention.
It is worthy to underline the contribution of the Aristotelian distinction between
good and bad habits, that intends good habits as those enhancing the agent’s control
to reach certain goals. Consequently, we can clarify the relation between habits and
emotions. The habits-as-learning entail control and for this reason they are funda-
mental to reach personal goals. This is the process that favors the agent’s pleasure
and happiness.
Some authors intend the idea of “habit learning” as the performance of an action,
previously learned after many repetitions namely in an unconscious manner, and
From Habits to We-Intentionality: Rituals as Social Habits 193
whose execution is inflexible and independent to the outcome [17]. This perspective
requires an integration with other perspectives that recognize the importance of the
sensitivity to the outcome and of different levels of flexibility and feedback [18].
According to Lombo and Giménez Amaya, a neurobiological view of “habit learning”
and recent experimental contributions (especially those of Graybiel) are consistent
with the Aristotelian notion of “habit”. Human habits are essentially based on two
aspects: (a) the stable character of an acquired quality; and (b) the capacity for new
actions that arises from that quality [19].
The integration between routines and goal-directed behavior is at the center of
important study in social psychology. Wood observes that successful long-term self-
regulation involves habitually engaging in actions that correspond with valued long-
term goals; people exert self-control to inhibit unwanted habits, and they repeat
goal-directed actions in stable contexts so as to form desired habits [20]:
Self-regulation of habits, even habits of addiction, may proceed as a result of people shifting
their goals and intentions to be consistent with the habits they perform. Although the inference
that a habitual behavior was intended is largely erroneous in terms of psychological processes,
these inferences have an intuitive plausibility for high-frequency behaviors. Such inferences
are further promoted by the switching costs of deviating, the fluency of habit performance,
and the potential benefits of habit performance for well-being.
movement level [22]. The researches indicate a close link between (socio)cognitive
abilities and situated action and try to enlighten their neural underpinnings and mech-
anisms. In this context, it has been revealed that the motor system is highly engaged in
anticipatory, simulative and generative processes. Some authors introduce an interest-
ing speculative perspective, and make the case that the same predictive mechanisms
provide both a ‘linkage with the future’ required for taking goal-directed action, and a
‘linkage with others’ required to act socially [23]. We observe that recent researches
have lead to a profound rethinking of basic concepts in cognitive and behavioral
sciences, and a common theoretical view—a motor-based (or action-based) view of
cognition—is emerging across disciplines [24]. They provide a description of the
abilities of action execution, its planning, and understanding of others’ intentions as
essentially goal-directed and served by the same representations, which are action-
oriented and involve deeply the motor apparatus.
Routines and goal-directed behavior characterize habits both in the case of indi-
vidual and social behavior. We can observe ritual behavior also in non-human ani-
mals, but, in the human case, they have their expression in a different form of We-
Intentionality. Moreover, they need conventional devices, that are mutually recog-
nized through ordinary communication. Rituals have the important function to create
social spaces in which individuals can share emotions, experiences, values, norms
and knowledge. The function to share experiences is fulfilled when there exist a social
space created by cooperation for reaching certain goal. In this context, cooperation
is a kind of intersubjectivity typical of human beings who, differently from apes,
are able to have “Collective Intentionality” [7] i.e. the basic intention to cooperate
and therefore to reach together a certain goal. If we want to get a positive result
about the extension of habits in the social dimension we need to move from a sort of
goal-directed activity that we can perform together. There is a contemporary lively
debate on the nature and structure of Collective Intentionality, as a necessary notion
to researches in the field of social ontology (the pioneers in this area are John Searle,
Raimo Tuomela, Margareth Gilbert, Michael Bratman and Philip Pettit).
We can observe that human beings (but also other species) have the capacity
to impose a function to an object so that the object acquires a function dependent
on the peculiar scope of the agent. The continuity between individual habits and
rituals (social habits) is thus showed by the fact that humans create these “agentive
functions” (in Searle’s terminology) in a wide variety of situations. Also non human
animals have their form of creating functions for objects but there is a fundamental
difference in the concept of “function” in the human case. I think that this point is
clearly explained by Searle and it is worthy to be quoted for our argument [25]:
(…) when we discover functions in nature, what we are doing is discovering how certain
causes operate to serve certain purposes where the notion of purpose is not intrinsic to mind-
independent nature, but is relative to our sets of values. So, we can discover that the heart
pumps blood, but when we say that the function of the heart is to pump blood, we take it for
granted that life, survival, and reproduction are positive values, and that the functioning of
biological organs sees these values (…) a function is a cause that serve a purpose. And the
purposes have to come from somewhere; in this case, they come from human beings. In this
sense, functions are intentionality-relative and therefore mind-dependent.
From Habits to We-Intentionality: Rituals as Social Habits 195
We create social habits in the form of rituals by using the “status function”,
which is a peculiar kind of function from which we create the social world. Rituals
are created by intentionality-relative functions and are characterized by two special
features. They require (a) a shared intentionality namely “collective intentionality”
and (b) collective imposition and recognition of a status.
The “constitutive rule” is essential to the process of constitution of institutions in
general. The canonical form introduced by Searle is:
Similarly, we can describe a very famous example of a ritual (Searle often refers
to), namely “marriage”. First, we need to be moved to act in a certain way. We-
Intentionality works when we want to do something together (we have a collective
intention) so that we can cooperate to achieve our common goal. As we already
anticipated, Collective Intentionality presents a weak form (collective recognition)
and a strong form (cooperation). Both are crucial for rituals, in our case marriage.
Now we can see how a social transition from one status to another is performed
through an institutionalized ritual:
(1) We have “collective recognition”, which means that the couple simply accepts
the institution of marriage prior to actually getting married.
(2) But, the actual marriage ceremony is an example of active cooperation, in
which the couple enters in a new social situation acquiring new social statuses
consequently.
(3) This fact obtains by the performance of the speech act of promise.
(4) The social context requires also the speech act of declaration from the part of
the institutional figure who has the suitable deontic powers to celebrate the rite
and to ascribe the new status to the couple.
5 Conclusion
We briefly introduced the notion of habit, which is at the basis of a lively con-
temporary debate. We think that a research that crosses philosophy and neuro-
science/neurobiology could offer some important clarification of the functioning
From Habits to We-Intentionality: Rituals as Social Habits 197
of habits and can extend to the social sphere of rituals and their function in individ-
ual and interpersonal contexts. Habits and rituals require routines and goal-directed
activity even though the context is different. Rituals requires also a kind of symbol-
ization that can be represented in different forms: the attribution of a symbolic value
to certain objects, animals and procedures. The object acquires a “status function”
i.e. counts as something that can be recognized to mean something else.
For instance, the ceremony to award diplomas requires the students to dress the
robe which means the passage to a higher level of education and a potential access
to a prestigious university.
In bullfight, people assign a symbolic value to the bull and to the peculiar uniform
of the toreador. Also in the case of rituals we can have “good” and “bad” practices
and the recognition of them implies the agents’ control on habitual behavior (habits-
as-learning) so that they can consciously choose to take part to ritual practices or to
change their choices. For example, to kill a bull is not considered a good practice
in every culture or for anyone (like eating lamb for Easter). Differently, to acquire a
higher level of education can be generally considered a good practice.
198 R. Giovagnoli
References
21. Castelfranchi, C., Pezzulo, G.: Thinking as the control of imagination: a conceptual framework
for goal-directed systems. Psychol. Res. 73(4), 559–577 (2009)
22. Iacoboni, M., et al.: Grasping the intentions of others with one’s own mirror neuron system.
PLoS Biol. (2005)
23. Pezzulo, G., et al. : The Challenge of Anticipation. Springer, Basel (2008); Pezzulo, G.,
Castelfranchi, C.: The symbol detachment problem. Cogn. Process. 8(2) (2007)
24. Jeannerod, M.: Motor Cognition: What Actions Tell the Self. Oxford University Press, Oxford
(2006)
25. Searle, J.: Making the Social World, p. 59. Oxford University Press, Oxford (2010)