Professional Documents
Culture Documents
Take Outs
Take Outs
888888888888888888888888888888888888888888888888888888888888888888888888888
constrained by the standards of the genre and some not easily articulated rules of how the
audience perceives music. The following is according to Max Martin, one of the top producing
I have lots of theories when it comes to [constructing a winning melody.] If you’ve got a
verse with a lot of rhythm, you want to pair it with something that doesn’t. Longer notes.
Something that might not start at the same beat… If the chords change a lot over the
course of a song, it’s better to stay within the same melodic structure. […] it’s all about
the balance. [reference]
[To be sure, there are infinite choices in how one can establish a correspondence between a
mental state and a musical structure. It seems describing the composer in the same way as we
describe a listener reducing possibilities of underdetermined content to one really belittles what
the composer does. But to say that the span of creativity is larger is not to say that the process is
not inherently the same.] Just as an audience in a conversation must reduce many possibilities
into a particular way of interpreting underdetermined content to one reality, the composer must
reduce many possibilities to make some arbitrary choices in how to interpret certain
The fact that the composer is forming a musical approximation of something he or she directly
something in the mind of another agent may raise a worry for how similar the two actions are.
But it is not clear how such worry translates into a concrete objection. Both cases amount to
attempts to take two states (in the first case mental state in the mind of composer and musical
2
state, in the second case mental state in the mind of speaker and mental in the mind of
instead. Whether the states are directly experienced may matter for how complicated of a task it
is but not for what kind of mental processes allow us to establish such correspondence. (If
anything, it seems that establishing a correspondence between direct mental states and musical
structures is more complicated than establishing a correspondence between mental states of two
There is also a worry that my portrayal of a composer as an interpreter of his or her own mental
states is a far departure from the Gricean program which so far has framed my discussion. After
all, communication for Grice involves getting an audience to recognize one’s intention to change
the audience’s mental state. If the process of converting mental states into musical structures is
at all like Gricean communication, what role do intentions play, who are they directed at, and in
virtue of whose recognition do they get fulfilled? Dan Harris defends a speculative hypothesis
that intentions do not have to be communicative to take advantage of the language system. He
gives an IBS account of speaking in order to clarify ones own thoughts by speculating that
intentions playing the role here might be “intentions to pretend to communicate or the intention
to hold something in memory for a short time.” I do not see a reason why an intention can also
be an artistic one, an intention to create poetry, to create music, to create art and artifacts. An
intention to change the mental state of another is just a type of an intention to change
something in general. Why does it have to be the mind of another? It could be to change the
Picturing a composer with intentions directed at him or herself and fulfilling them in virtue of his
or her own recognition seems like an unnecessary complication to describe how one converts
mental states into music. And yet, […]. So is this a problem for my argument or is it a problem
for Gricean program? I would not be the first one to say that there is something problematic in
[…]. Dan Harris has a discussion of how language use not involving an actual addressee
presents a problem for Grice. [The idea that we use language to clarify our thoughts seems to
entail that the content of one’s utterance is sometimes more articulate than the intention one has
in expressing it. But how could this be compatible with the claim that what one says is fully
determined by what one intends in saying it? And we seem to be put into a vicious circle by the
claim that the meanings of natural-language expressions can be explained in terms of the
The most glaring departure from the Gricean program in how I describe a composer as an
interpreter of his or her own mental states is that there is an intention to affect an audience and
that the intention is fulfilled in virtue of being recognized. [An objection about directing
intention. ok so, if not constraint maybe vector matters – problems with that. Or intention. ] But
[blah blha blha] is a glaring departure from the Gricean program that so far has framed my
There
Many artistic creations are links in the chains, taking the meaning created by some as inputs but
also serving as inputs in the creative expressions of others. A musician interprets the composer
score, creating musing which is then interpreted by a dancer, interpreted by an artist in creating a
Things do not stand differently when we consider the distinction between meaning and
interpretation in a regular conversation, albeit in a less obvious way. A large part of our day to
day communication is relaying: we relay to each other what we have learned from others, what
has happened to us, and how we feel. In such relaying, we both interpret other people's
messages, events, and feelings, and re-convey our interpretations to others with our own agendas
in mind. Just as a conductor cannot be viewed as a mere middleman between the composer and
the audience, no interlocutor should be viewed as lacking unique preferences and purposes in
relaying the information he or she has gained somewhere else. As interpreters, we are also never
impartial or un-creative. No communication will fully determine how it will be perceived: just as
the musical score leaves room for creativity of the performer, any communication will leave
choices for the interpreter to settle on his or her own, according to his own background, agendas,
and creativity.
There is an even more direct way of seeing how each act of meaning and each act of interpreting
is in reality both, depending on the perspective one adopts. When a speaker utters words directed
at an addressee, he or she is also engaged in interpreting his or her own state of mind. Just as
the musical score does not fully determine how the music should be played, the thoughts,
5
intentions, and some kind of utterance planning mechanisms that direct our speech do not fully
determine the words that the speaker will utter. There are some speculative hypotheses why that
may be, but the phenomenon of clarifying out thoughts through speech is widely accepted. We
sometimes even surprise ourselves with new understandings as results of our own utterance.
(There is a charming quote from a famous philosopher Bertrand Russel about confessing his love
to Lady Ottoline: "I did not know I loved your till I heard myself telling you so -- for one instant
I thought "Good God, what have I said?" and then I knew it was the truth.")
All participants in this chain perform what can be described as creating content (or conveying
meaning) and consuming content (or deriving an interpretation), and if we try to draw a border to
decide how much of one or the other is allowed in order for the person to still act under There is
no way to draw a non-arbitrary border between content-creating processes and interpretive ones
All of these roles under certain descriptions appear the composer’s role is not purely creative
either and requires some amount of conformity and guessing that we typically associate only
with the interpreter’s role. And if we try to figure out how much of this “leakage” or “blurring”
of the roles is allowed for them still to remain conceptually separate, we will quickly run into the
problem that the whole exercise is just arbitrary drawing of the borders.
Articulating what the objects of our beliefs are is so difficult because we try to describe an object
in a different medium in the medium of language – it’s like trying to describe poetry in regular
language. The objects in the medium of thought do not have a one to one correspondence to the
Underdetermination arises from the fact that the objects in different mediums do not correspond
to each other on the one to one basis. You cannot translate from one medium into another and
get an exact match. The more different are the mediums, the more underdetermination. There is
less underdetermination in translating poetry into regular language (at least you can capture most
of the meaning), but there is more underdetermination when you translate from music to dance.
[But what about from musical notation to music? Maybe it’s not different mediums? The
Maybe the problem with AI understanding human language is that creativity aspect is missing,
the ability to fill out the underdetermined content. You can run neural networks all you want, but
all they do is record patterns that are already there, they do not create anything new. I think
people do. How does that kind of creativity work? A lot has been written on the interpretive
process, but I am not familiar with any literature on creativity process. And since I think they are
the same, I also think the problem with everything written on interpreting is missing the
creativity part.
Grice was possible one of the first who noticed that an producer and consumer roles were
somewhat blurred – that the interpreter almost always has to create content in order to interpret
and that a speaker does not start with a blank page but is rather constrained in the kind of content
he may produce. What we have not noticed before is that this is not simply a blurring, but rather
the fact that the roles are inherently the same. One thing that still may seem to require a
consumer producer designation is the vector of the content – who it is derected at. However, the
7
direction of the content does not make the process of its production inherently different from the
process of its interpretation. Also, as the conductor example shows, a vector is also sometimes a
relative thing. In another example, I may be partially speaking to an interlocutor, and partially
musing to myself.
Can we object by saying that the conductor is a sort of pass through of the original content to
other interpreters, albeit with some creative additions? In other words, we may resort for a
The conductor is still acting primarily an interpreter, mostly performing all the associated
interpretive mind processes, but whenever he or she comes across some indeterminacies of the
score then some creative mind processes kick in. However, a description of a conductor as a
It strikes me that the phenomenon of openness you describe is exactly the phenomenon that
arises in communication between an artist and the audience, a composer and a performer, a novel
writer and a reader, and those examples might be useful in figuring out what is going on in a
instead, it is a pervasive feature of how we communicate in all kinds of ways. In the case of art,
when such indeterminacy arises, it does not seem surprising to us -- of course, there are many
legitimate ways to interpret a novel or perform a musical score! Why is it surprising to us that
there is also openness to interpreting a conversational interlocutor? We don't even realize that the
assumption that it is possible is in the back of our minds, usually, when confronted, we would
I had a few observations, that I was wondering if I could get your thoughts on...
Buchanan's way of dealing with the "openness" of implicature is to say that an implicature is not
communicating any specific proposition, but rather a type of a proposition. But something in this
proposal seems unsatisfying to me. When we look at some possible propositions within the type
that is supposedly being communicated, we see that the only feature that they share is that they
pass the filter of the constraints that the speaker imposes on the possible interpretations. Those
constraints are the linguistic meaning of the words, but also the shared knowledge that
handwriting and punctuality do not bear on someone's qualifications as a job candidate. The
endless variety of propositions that fit the type share only one quality, and that's the quality of
passing the filter of those constraints. But to designate them as a type seems tautaulogical for
that reason seems kid of tautaulogical: what is being communicated? whatever fits the
constraints? what fits the constraints? a type of proposition. what type is that? one that fits the
constraints.
translate content from one medium to another: from the musical score to a musical performance,
from musical performance to the movements of a ballet dancer, from the architect's drawing to a
building, from the from the linguistic meaning of a novel to an interpretation in the mind of a
reader. In all these cases, the medium of the first member of the pair has only vague
correspondence to the medium of the second member, not one to one. There are many possible
ways to perform the same musical score or choreograph the same ballet music because the
What is the difference in mediums that produces indeterminacy in the case of Smith? Smith is
clearly trying to cause the listener not to hire the student, to cause a decision in the listener's
mind to not hire the student. But what is the medium of a decision? A greatly controversial
topic. But most probably would agree that at least not all thoughts are in the form of propositions
and not in the forms of types of propositions. For example, the decision to eat an ice-cream
probably does not come about as a proposition "I am going to eat an ice-cream" and not even as a
type of a proposition about eating ice-cream. I am not going to speculate what medium decisions
are in, but if Smith is trying to get the listener to decide not to hire the student, that kind of
When we say that content is specific -- it seems that it is expressed in the same medium, and
A different model of communication. We don't send each other parcels of content. We send each
There are a couple of options. Say, the output of the interpretive process includes the making the
creative choices to fill in what is unspecified by the score (the exact tempo, the decision to
emphasize a certain part, the character of certain sounds, etc.). In this case, the task is already
exhausted before content production even begins, since there is no other kind of content that the
conductor can create while still adhering to the constraints of the original score. Now suppose
the output of the interpretive process pertains only the specified content and serves as an input
10
for choosing how to fill in the unspecified content. For example, specifies which note
to play but does not specify the duration. Can we say that the former kind of content serves as an
producing content or interpreting it cannot hinge on an arbitrary decision exactly how much
constraint or how much creative choices is appropriate for that to be counted as one or the other
role. ]
http://www.sea-acustica.es/fileadmin/publicaciones/Sevilla02_mus01008.pdf
This is an important note, to which I will return: how the necessity to supply the unspecified
content arises only when we translate content into a different medium.] Identifying the
constraints and filling in specificities is the same process which we sometimes describe as
content production and sometimes content interpretation. But the best argument for this is that
[How do I reconcile this with neuro data, how people can lose an ability to speak but not
There is an idea among communication theorists (and we already came in contact with it) that
some content is precisely specified by the medium of communication while some content is
“plays the orchestra”). There is never a single way of realizing the same instruction, however
precisely stated, in the medium of sound. Say, the pitch conveyed by a note is a precise enough
instruction, while the duration is not quite. There is no way to realize the pitch of a particular
note without also realizing its duration. To say that realizing the instructions given by the same
musical note requires two co-occurring but inherently different mental processes – one
[more on how inferential thinking is constrained thinking. Boghosian thinks it’s rule following
thinking. It seems to involve at least identification of some constraints. While creative thinking
https://www.frontiersin.org/research-topics/941/expression-of-emotion-in-music-and-vocal-
communication#articles
[It seems that something like penumbral shifts happens in how we use physical artifacts. We are
certainly not as mystified by it when it happens with physical stuff. Take a token of an arrow
used for hunting. It can be successfully used to kill a rabbit. However, in a different context,
one that requires taking down a mammoth, the hunter will say, “No, this is actually not going to
work here. Not quite sharp or sturdy enough,” and pick a much larger token or another tool
altogether. A hunter may similarly struggle to articulate the exact contexts in which this token of
12
arrow will serve well and in which it won’t, very similarly to how a speaker may struggle
articulating the exact rules of where the word “boy” applies, but that will not prevent the hunter
from using the arrow successfully just as it won’t prevent a speaker from using the word
successfully. We often demonstrate the same penumbral ignorance and indifference about many
change the mind of the other, in using artifacts, we have intentions to change the environment
around us (an intention to kill a rabbit, to flip a pancake, to build a house), and in both cases, we
search for the best tools to fulfill those intentions. In the latter case, we are constrained by the
environmental context (how small and soft is the rabbit) and by the characteristics of the tools
(how sturdy and sharp is the arrow). We never know for sure if the tool will be good enough to
fulfil a particular intention – we always run the risk that the arrow will be just a little bit too dull
to kill the rabbit – but we manage, for the most part. Our purposes are often novel, and we
manage either with old tools or by modifying them and inventing new ones. More often than
not, our purposes are so complicated that require multiple steps with multiple tools. None of this
seems to require anything like “compositional semantics” of how we manage to fulfill our
infinite purposes with our tools. Natural language is also often thought of as a toolbox. It is also
certainly not static – old tools get modified all the time, and new ones get added in depending on
our needs. What is different about language that presses us to search for something like a
Are the parallels I am trying to construct how we use language expressions and physical tools too
crude? You can work out parallels to linguistic meaning (generally known purposes of tool-
types: “a chair is for sitting”), content (what a particular tool is being used for on a particular
occasion), and the parallels to being true or false is simply being successful at fulfilling a
13
particular purpose. We know that chairs are mostly for sitting, but there are many other uses of a
chair that would not surprise us that do not constitute sitting (resting our feet on, leaning on,
putting our stuff on) and there are many chairs that may not fulfill the purpose of sitting in a
particular situation (too small, too weak, etc). This seems to me a lot like vagueness of
expressions.
Can we distinguish what instructions the conductor or an independent player is merely inferring
and what instructions they are producing? If it is a performance of a particular scored work, all
of the content – whether the kind that is inferred or the kind newly created – must be constrained
by the composer’s score, various kinds of knowledge shared between the composer and the
performer, and the principles of musical cooperation. Therefore, the only way a performer can
create new content is by choosing from a pre-existing range of possibilities. Could we then say
that whenever the performer is inferring what is precisely specified, then the performer is acting
as a consumer, and whenever the performer is choosing from a range of possibilities conforming
Even the most subtle of possibilities present opportunities for the performers to create their own
Different orchestras will respond differently to the same conductor, as Zubin Mehta illustrates:
"If I give the first beat and nothing happens, [I know that I am standing in front of the Vienna
Philharmonic]." A recent study has confirmed that different orchestras play with distinctive
14
styles, recognizable by professionals and non-musicians, which cannot be distilled to the efforts
of a conductor [reference].
description of the audience is missing for whom it is creating. But is this objection against
me or against Grice? Latter, because similar worries have been raised against Grice: use
of language for inner speech seems meaningful but it is missing for whom it is creating. I
am hoping that my examples actually help IBS respond to these objections. This shows
that presence or absence of an audience should not matter for whether we describe
departing from Gricean requirement that to mean one strives to change the mind of the
other.]
as constructing a musical approximating for a mental state the composer experiences. Just as an
meanings and their utterances to form an approximation of what a speaker means on a given
occasion, a composer may exploit common correlations between mental states and musical
structures to form a musical approximation of a mental state. The correlations between musical
structural features (such as tempo, mode, volume, melody, and rhythm) and mental states (such
as feelings of joy and sadness, and even propositional attitudes, such as hope and fear) have been
a subject of extensive empirical research [reference]. While the source of such correlations is
15
controversial (whether they are conventional or have some deeper neurological causes), their
existence is not.