Compositionality: A connectionist variation on a classical theme

Gelder, Tim van

doi:10.1016/0364-0213(90)90017-q

Cited by 235 publications

(62 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…I know only two clear accounts of this mereological relation-the language-of-thought hypothesis developed, e.g., by Fodor (1975) and Van Gelder's (1990) attempt to develop a notion of constituency compatible with connectionism (following Smolensky 1989; see also Smolensky and Legendre 2006). It is unclear whether Hill and others endorse any of these accounts.…”

Section: Thus C Is Incorrectmentioning

confidence: 97%

Replies to my critics

Machery

2010

Philos Stud

View full text Add to dashboard Cite

Section: Thus C Is Incorrectmentioning

confidence: 97%

Replies to my critics

Machery

2010

Philos Stud

View full text Add to dashboard Cite

“…Regarding the third premise: Chemero mentions a number of connectionist (e.g., van Gelder 1990 andChalmers 1990) who argue that connectionist networks do have representations with compositional structure. So Fodor's and Pylyshyn's third premise rules out a priori connectionist networks with compositional representations.…”

Section: Identifying the Question-begging Argumentmentioning

confidence: 99%

The coupling-constitution fallacy

Kagan

Lassiter²

2013

P&C

View full text Add to dashboard Cite

The coupling-constitution fallacy claims that arguments for extended cognition involve the inference of "x and y constitute z" from "x is coupled to y" and that such inferences are fallacious. We argue that the coupling-constitution fallacy fails in its goal to undermine the hypothesis of extended cognition: appeal to the coupling-constitution fallacy to rule out possible empirical counterexam ples to intracranialism is fallacious. We demonstrate that appeals to coupling-constitution worries are problematic by constructing the fallacious argument against the hypothesis of extended cognition. We consider several objections to our argument and find them insufficient to rebut our conclusion.

show abstract

“…But, as Everett notes (p. 94), at least one of the questions to ask is "Is there a simpler explanation?" This has led to many attempts to model the systematicity of language on the basis of a representational system without constituent structure (e.g., distributed representations utilized by connectionist networks; van Gelder 1990), that have, however, encountered significant difficulties (e.g., Hadley 1994;Haselager and van Rappard 1998). One recent attempt, that I think is congenial to Everett's perspective (e.g., pp.…”

Section: Systematicity and The Breathability Of Airmentioning

confidence: 99%

Like the breathability of air

Haselager

2012

P&C

View full text Add to dashboard Cite

I present experimental and computational research, inspired by the perspective of Embodied Embedded Cognition, concerning various aspects of language as supporting Everett's interactionist view of language. Based on earlier and ongoing work, I briefly illustrate the contribution of the environment to the systematicity displayed in linguistic performance, the importance of joint attention for the development of a shared vocabulary, the role of (limited) traveling for language diversification, the function of perspective taking in social communication, and the bodily nature of understanding of meaning.

show abstract

Compositionality: A connectionist variation on a classical theme

Cited by 235 publications

References 4 publications

Replies to my critics

Replies to my critics

The coupling-constitution fallacy

Like the breathability of air

Contact Info

Product

Resources

About