Desktop version

Home arrow History

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Enter roots

The nature of roots—objects which are manipulated by syntactic processes and which are supposed to provide access to conceptual system(s) of the mind—is one of hotly debated issues in recent theorizing about the relationships between the faculty of language in the narrow sense and its surroundings. On a conservative stance, a root is associated with a complex lexical meaning treated for the purposes of syntax as an atom—a mere pointer to a concept lying outside the bounds of grammar and inaccessible for syntactic processes—as it may be expected to be given the basic architectural assumptions of the generative theory and the thesis of the autonomy of the syntactic module; where approaches differ is first and foremost the answer to the question how roots are related to concepts exactly and which properties, if any, of their conceptual counterparts may be inherited as syntactically visible properties—are they entirely ‘bare, deprived of features which affect their syntactic behaviour (e. g. De Belder and van Craenenbroeck (2015) or Marantz (2007): distinct as they are, both kinds of approach to roots deny that roots have inherently properties determining their argument-taking behaviour), or do they have properties like adicity (as e. g. on the proposal of Harley (2014))? Or, perhaps, although related to concepts, are they endowed with different argument-taking properties than their conceptual mates (as e. g. in Lohndal and Pietroski (2011), Pietroski (2014a) and related work)? An implicitly assumed traditional stance on the relationship between conceptual resources and lexical items entering the derivational space would link the two components in a direct way:

One familiar suggestion is that lexical items simply label the concepts they lexicalize, and that composition of lexical meanings mirrors composition of the labeled concepts, which exhibit diverse adicities. This makes it tempting to say that names label singular concepts—mental tags for particular things—which can saturate the concepts labeled with verbs, where these predicative (unsaturated) concepts may be monadic or polyadic. (Pietroski 2012b: 129)

Such picture would then significantly reduce the extent to which lexicalization, understood as a process which enables concepts to enter into syntactic configurations, would be an innovative process, making it essentially a process of translation from one language into another. It would still be a cognitively significant achievement if the translation involved more than one source language, thereby allowing concepts originally belonging to distinct modules to form together the set of building blocks for the syntactic engine—the cross-modular nature of human thought is repeatedly stressed in much research (see e. g. Reinhart (2006), Spelke (2003), Carruthers (2006), a. o.). It would be also still compatible with the assumption that the innovation consists in adding an ‘edge feature’ (Boeckx 2011a,c, 2012a,c), understood in the manner of Pietroski (2012b) and Boeckx (2015a) as ‘an instruction to "fetch a concept C’”, so that for two concepts C and C' which ‘may not on their own be readily combinable, owing to their belonging to different cognitive modules (by definition, the combination, or merger, of modular concepts is restricted, or ‘encapsulated’), the merging of {C} and {C } results in cross-modular union' (Boeckx 2015a: 101-102) (which is not to say that Pietroski (2012b) or Boeckx (2015a) subscribe to this scenario—neither does, and both stress the fact that combinatorial abilities of lexical items would be in such a case constrained by properties of original concepts; still, the cross-modularity effect would be achieved, although within bounds set by adicities established in distinct modules for different concepts). Furthermore, it would be still compatible with the crucial property of the transition from perception-related modules to the linguistic faculty—the change from being stimulus-bound to being part and parcel of the spontaneous activity of the mind:

The process of lexicalization de-couples the percepts that are selected from their respective visual stimuli, giving us new and more abstract entities, lexical items, which are stimulus-free and independently manipulable, enabling creative thought and reference.

For this reason, we may describe this process as one of ‘de-indexi-calization’. (Hinzen and Sheehan 2013: 47)

Yet the fact that, as currently assumed, a part of the conceptual system in the sense required here was recruited for the language faculty in the narrow sense, another part coming to being only with the presence of the latter—as Chomsky (2009: 29) puts it, ‘emergence of unbounded Merge in human evolutionary history provides what has been called a “language of thought,” an internal generative system that constructs thoughts of arbitrary richness and complexity, exploiting conceptual resources that are already available or may develop with the availability of structured expressions’—suggests that this picture might not do justice to the import of the transition from an independently available conceptual repository to a reservoir of atoms of syntactic computation. On the analysis put forward in Pietroski (2014a) the transition involves turning a concept into a (counterpart of) a monadic predicate, conjoined further with other monadic predicates in the C-I component, with a crucial ingredient being availability of predicates expressing 0-roles for the interpretation to proceed; while not being the only game in town, it highlights problems arising in the transition from the repository of pre-linguistic concepts to the lexicon, problems going beyond mere (in)ability to undergo merge, which manifest themselves particularly clearly in the case of relation concepts.

<<   CONTENTS   >>

Related topics