The School of Advanced Studies

Professor 

Wolfram Hinzen
ICREA/Universitat de Barcelona

Number of hours: 3.5h hours total

 Date:
3/17/15

Schedule timetable: Tuesday  3/17/15 from 09am to 10h45am
                                    Tuesday  3/17/15 from 11am to 12h45am

Description : 

      Classical generative grammar, conceived as a form of ‘Cartesian Linguistics’ (Chomsky, 1966) methodologically abstracted from the connection that language has with thought, studying it formally as a separate system in its own right. This didn’t prevent matters of content from arising, and recently a view has gained momentum according to which the generative system underlying human language is the generative system underlying human thought, insofar as it is species-specific. Yet how can the study of grammar be the study of a thought system, and provide a theory of thought?

    It can if Homo sapiens engages in a peculiar form of conscious thought that is uniquely linguistic, with the basic organizational principles of grammar determining those of this mode of thought, which we do not find in non-grammatical species. Whatever we think of ‘non-linguistic thought’, and whatever ‘language of thought’ exists without ‘real’ language, such thought is different. Language might make this difference. But how? Grammar might give our minds meanings to think that would not exist without it, and empirically do not seem to exist without it. In these two sessions I introduce this new concept of ‘Un-Cartesian Linguistics’ to shed light on how defining features of human-specific thought can actually fall out from grammatical organization, with a focus on how this is true for our sense of selfhood.

       Selfhood has been taken as constitutive for any form of rational thought since Kant, who maintained that all thoughts must be subordinated to the ‘Ich denke…’ (I think that…). Along with Kant, a long tradition has taken personal-pronominal forms of self-reference to be essential and irreducible to one kind of selfhood as such. This invites renewed reflection on the cognitive significance of grammar, and it makes predictions for the linguistic profile of patients with problems of mental health that present with disturbances of the self.  

     In line with these predictions, the pronominal system is specifically affected in cognitive disorders such as autism and schizophrenia, and I will develop this point into a more systematic assessment of the role of language in these cognitive disorders, with a focus on schizophrenia. Language disturbances lie at the core of the psychopathology involved, and in turn, for almost a century schizophrenia has been described as a disturbance of selfhood (ipseity). The question is how these two facts relate.

Chomsky, N. 1966. Cartesian Linguistics. New York: Harper & Row.

Hinzen, W. & M. Sheehan 2013. The philosophy of universal grammar. Oxford: Oxford University Press.

Martín, T. and W. Hinzen 2014. The grammar of the essential indexical. Lingua 148, 95—117

Professor 

Anna Maria Di Sciullo
UQAM - Canada    

Number of hours: 3.5h hours total

 Date:
3/18/15 & 3/19/15

Schedule timetable: Wednesday 3/18/15 from 9am to 10h45am
                                 Thursday    3/19/15 from 9am to 10h45am

Description : 

         The growth of language in the individual is determined by genetics, experience and principles of computational efficiency. The latter are taken to be part of natural laws affecting the development of biological systems. I discuss the effect of two principles of efficient computation applying to the derivation of linguistic expressions and their interface representations. I develop the hypothesis that these principles contribute to language variation and acquisition, given the domain specific properties of the human language faculty. In this perspective, I contrast language and mathematics. I focus on Indirect Recursion (IR), the recursive merger of a given projection X mediated by a functional element F: [X [ F X ]]. I posit that IR is forced by the principle of efficient computation Minimize Symmetrical Relations (MSR), whereas it is not necessarily legible at the sensori-motor (SM) interface as enforced by the principle of Minimize Externalization (ME).

       I discuss the results of neuro and psycholinguistic studies on the processing of complex nominals in Romance language and in English, which bring experimental support to my hypothesis. In conjunction, these principles provide an account for the variation and the apparent gradual development of functional categories in the individual. Furthermore, I provide evidence that IR, enforced by MSR and ME, hold for complex numerals, according to language specific parameters, differentiating Russian from Arabic, for example. I discuss recent contributions of neuroscience in the identification of species-specific brain pathways for language and mathematical computations. IR is generated by the computational procedure of the human language faculty, while concatenation is available for mathematical operations, in humans and animals. MSR and ME affect the computations in the dedicated regions of the human brain activated in language, predominantly BA 44-45; whereas such principles do not affect the processing of mathematical formulae, which strongly recruits a more anteriority located region, predominantly BA 47.

        Theoretical and experimental results indicate that the MSR and ME principles affect the computation of complex nominals and complex numerals by the human brain; whereas there is no evidence that this would be the case for mathematical formulae.

Professor 

Anne Christophe
ENS - PSL Research University

Number of hours: 3.5h hours total

Schedule timetable: Wednesday 3/18/15 from 11am to 12h45am
                                 Thursday    3/19/15 from 11am to 12h45am

Description : 

        For a long time, children were thought to acquire first the sounds of their native language (its phonology), then its words (or lexicon), then the way in which words are organized into sentences (its syntax). This corresponds to what young children produce: first they babble (between 6 and 12 months), then they speak in isolated words (1-2 years), and then they start combining words together. Accordingly, researchers have looked for ways in which children may acquire the sound system of their language before they know words, words before they know syntax, and so on. In many cases however, computational studies have shown that some learning problems are intractable unless one postulates access to at least partial information from other domains, and experimental studies have shown that children have managed to learn some of this partial information. I will present experimental and computational work that tackle acquisition problems where synergies between domains have been demonstrated (between phonology and lexicon, and between lexicon and syntax). 

Professor 

Robert Berwick
MIT USA

Number of hours: 3.5h hours total

Schedule timetable: Friday 3/20/15 from 9am to 10h45am
                                   Monday 3/23/15 from 9am to 10h45am

Description : 

        The evolution of the faculty of language largely remains an enigma. In this course, we ask why. Language's evolutionary analysis is complicated because it has no equivalent in any nonhuman species. There is also no consensus regarding the essential nature of the language “phenotype.” According to the “Strong Minimalist Thesis,” the key distinguishing feature of language (and what evolutionary theory must explain) is hierarchical syntactic structure. The faculty of language is likely to have emerged quite recently in evolutionary terms, some 70,000–100,000 years ago, and does not seem to have undergone modification since then, though individual languages do of course change over time, operating within this basic framework. The recent emergence of language and its stability are both consistent with the Strong Minimalist Thesis, which has at its core a single repeatable operation that takes exactly two syntactic elements a and b and assembles them to form the set {a, b}.

Professor 

Randy Gallistel
RuCCS, USA

Number of hours: 3.5h hours total

Schedule timetable: Friday 3/20/15 from 11am to 12h45am
                                 Monday 3/23/15 from 11am to 12h45am

Description : 

       From the traditional perspective of associative learning theory, the hypothesis linking modifications of synaptic transmission to learning and memory is plausible. It is less so from an information-processing perspective, in which learning is mediated by computations that make implicit commitments to physical and mathematical principles governing the domains where domain-specific cognitive mechanisms operate. We compare the properties of associative learning and memory to the properties of long-term potentiation, concluding that the properties of the latter do not explain the fundamental properties of the former.In this course I will briefly review the neuroscience of reinforcement learning, emphasizing the representational implications of the neuroscientific findings. I will then review more extensively findings that confirm the existence of complex computations in three information-processing domains: probabilistic inference, the representation of uncertainty, and the representation of space. I argue for a change in the conceptual framework within which neuroscientists approach the study of learning mechanisms in the brain.

Professor 

Colin Phillips
U. Maryland

Number of hours: 3.5h hours total

Description : 

        Linguists are impressed by the rich grammatical details that natural languages follow. There is now abundant evidence that speakers and comprehenders show fine-grained control over these details during moment-by-moment speaking and understanding, but how do they do this? To make matters more interesting, much recent research provides compelling evidence that language users make use of domain-general memory access mechanisms to retrieve words and phrases and to form linguistic dependencies during comprehension. But these domain-general mechanisms, which access information based primarily on content, are not straightforwardly compatible with pervasive constraints that focus primarily on structural configurations. I will discuss the memory mechanisms, the linguistic constraints, the current evidence on how to reconcile them, and key questions for future research.