Professional Documents
Culture Documents
Basel
Human Development 1990;33:362-365 001 8-716X/90/0336-0362S2.75/0
be tampered with in its present form. He cri cates like Pylyshyn and Fodor on the propo
ticizes my claim that logics that are shown to sitional (logical) side of the debate and Shep
Downloaded by:
UCL
Piaget's Theory' 363
ard. Kosslvn. and others on the analogical ments.’ His theory is not concerned with the
side [Pylyshyn. 1981], Even earlier, serious development of structures, but with so-
debate over the type of analogical model called structure-mappings and their develop
Halford proposes occurred between Hutten- ment. Structure-mappings are rules for ‘as
locher and Clark [Huttenlocher et al., 1972]. signing elements of one structure to elements
Some argue that the issues, or merits of one of another structure’ [Halford. 1988. p. 104],
side over the other, are undecidable. Thus, it Even if Halford defined structure as does
is more than premature for Halford to assert Piaget, which 1 do not believe he does, the
that the claims for logical models of reason emphasis in his theory is on functional
ing have been seriously faulted and that one change and on relations that are more se
should look to analogical models as a more mantic than functional. Even his definition
viable alternative. My position is that we do of structure is based on functional features,
not know enough about logical, linguistic or rather than on holistic properties of struc
analogical models to categorically reject any tures and the possible transformations
one of them [Beilin, 1975]. Rather, the evi among them.
dence in some sense supports each of the The issues are complicated by the incor
models, and in my view a model that incor poration of elements from structuralist the
porates both functional and structural forms ory into contemporary information-process
of reasoning appears the most likely solution ing theories [Beilin, 1983]. These structures
to the issues posed in these debates. take the form of semantic networks, decision
There is much in Halford’s present rejoin trees, lists, and other entities, for the organi
der about structure. Would one on the basis zation of knowledge in memory. As a rule
of his comments count him among structur these theories, again, are not concerned with
alists? He says. ‘1 certainly agree that all cog system or other holistic properties of struc
nitive architectures entail structure in some tures, logical or otherwise, but with the func
sense’ [Halford. 1990, p. 360]. In some tional properties of systems, and. borrowing
sense, Skinner, Watson and Pavlov would from earlier functionalist theories, they em
count as structuralists, in that connections, phasize associational and dimensional rela
bonds or associations between stimuli and tions. Neo-Piagetian theorists, in contrast,
responses, or responses and responses, can retain more features of structuralist think
be considered structures. Halford [1990, ing, to varying degrees - Case to a greater
p. 357] is more decisive, ’... my position has degree, Halford and others to lesser degrees
always been explicity structural’. But. he says [Demetriou. 1988], The critical question, as
further, i view structure differently from the I see it. is to what extent developmental the
way Piaget viewed it.... I postulate mental ories can be successful if they ignore the
models that represent the environment. structural features of thought, and, if they do
They are actively constructed, and the con not. as is the case with neo-Piagetian theo
struction is guided by experience’ [Halford, ries, what forms those structures need to
1990. p. 357], He also says in a recent paper take, and what assumptions are made with
[Halford. 1988. p. 104]: ‘We will define a respect to their nature.
144.82.238.225 - 1/11/2018 5:32:36 PM
structure as a set of elements, with a set of Halford characterizes his mental models
relations and functions defined on the ele as representational, i.e.. they represent the
Downloaded by:
UCL
364 Beilin
environment. Piaget’s structures are not re theory, in his view, provides a better expla
presentations of the environment or any nation than Piaget’s. Further, Halford’s
thing else: his representations are semiotic statement about Piaget's theory to the effect
instruments of thought. In characterizing that ‘... children cannot perform certain
Halford's theory as based on changes in re tasks because they have not entered a partic
presentations, I did not mean that he had ular stage is not a satisfactory explanation, as
adopted Bruner's representational theory, it only summarizes the data’ [1989, pp. 351 —
but that his theory was analogous to Bruner’s 352] is one of the reasons I called Halford’s
in basing cognitive development on changes view of Piaget’s theory ‘overly simple’ (if not
in representations, rather than on changes in simplistic), considering that Piaget’s vast
structure. My point is that Halford's theory body of research and theory provides an
rests on functional representations of a di astonishingly rich account of basic mecha
mensional character and not holistic, sys nisms in development.
temic structures that tie distinctive constitu It is salutary to have Halford’s assurance
ent properties together into coherent units that in a 1982 publication he ‘recognizes that
having properties that transcend those of its developing systems can be both continuous
constituents. Halford's definition of struc and discontinuous’. Unfortunately, neither
ture departs from Piaget’s, and does not in the 1989 paper nor in the present rejoin
merely differ from it in content. I would like der does he indicate what he believes is con
it understood that I am not arguing against tinuous and what is discontinuous in a de
models that place their major emphasis on veloping system.
functional or process aspects of develop In conclusion, as I observed in my origi
ment, as do information-processing models nal critique of Halford’s paper, he does a fine
for the most part, nor against neo-Piagetian job of evaluating the merits of various tests
models of which Halford’s is a distinctive of Piaget’s theory. His evaluations are not
example. My own disposition, however, is always complete or well-balanced, but on the
toward models that integrate more truly whole, they explore important issues in a
structural forms with the functional, as the highly complex and extensive literature and
later Piagetian theory attempts. the evaluations are often sensitive and so
Halford asserts that Piaget’s theory is ‘at phisticated.
best descriptive' and not explanatory. This is Halford may be correct in believing that
an old claim, and was extensively debated by insufficient attention has been paid to test
Brainerd [1978] and a group of discussants, ing neo-Piagetian theories of cognitive de
so there is no need to repeat that debate here. velopment. including his own. It is not, how
Piaget [1985/1975] himself deals with the ever, a constructive contribution to the de
same question. Unless Halford can come bate on the issues to make a tendentious
forth with a new view of the nature and role claim that all Piaget’s theory merits is ‘ceas
of explanation in science, we have to take his ing trying to test it’. This is particularly the
claim that ‘analogical structure mapping’ case now when a large new body of Piaget's
can explain difficulties in reasoning while research and theory is still appearing, re
Piaget’s theory cannot, as vacuous. A less search that offers new and fresh insights into
144.82.238.225 - 1/11/2018 5:32:36 PM