KogWis 2012

Symposium Linguistic vs. Non-Linguistic Knowledge

30 September - 3 October 2012
University of Bamberg

Barbara Kaup

Psychologisches Institut

Universität Tübingen

barbara.kaupspam prevention@uni-tuebingen.de


Claudia Maienborn

Deutsches Seminar

Universität Tübingen

claudia.maienbornspam prevention@uni-tuebingen.de

For understanding language we may exploit a huge variety of rich knowledge resources: linguistic structure, contextual information, sensory-motor input, world knowledge, pragmatic strategies, etc. The question of whether there is a need for systematically distinguishing between linguistic and non-linguistic knowledge is of interest for both linguistics and cognitive psychology. Yet, until recently both sides tended to ignore this issue more or less. Theoretical linguists usually suppose tacitly that there is such a difference but then only take care of (what they assume to be) the linguistic meaning part. (A prominent exception is Asher’s (2011) theory of lexical meaning in context.) In cognitive psychology the standard assumption is that word forms gain meaning by being connected to nodes in the conceptual system. However, the exact nature of this relationship is rarely discussed. With a few exceptions (e.g., Vigliocco & Vinson, 2007), most authors seem to assume that word meanings and concepts are identical.

This issue is particularly relevant because it points to the status of compositionality as a guiding principle for the formation of complex meanings. In a frequently cited study by Hagoort et al. (2004) the distinction between linguistic and non-linguistic knowledge has been challenged by experimental results. Hagoort et al. concluded that semantic and conceptual information is processed in one step, making obsolete the distinction, and challenging, furthermore, the idea of compositionality which proc laims the primacy of syntax as a combinatorial guideline. This controversial issue has been taken up more recently in several experimental studies (e.g., Pylkkänen et al. 2009). What can be learnt from these studies is above all, that we are strikingly missing a good understanding of what could count as methodically and empirically solid criteria to distinguish linguistic and non-linguistic knowledge. The symposium will provide an interdisciplinary platform for linguists and cognitive psychologists to discuss questions pertaining to the distinction of linguistic vs. non-linguistic knowledge.


Linguistic structure meets natural language metaphysics
(Nicholas Asher, CNRS Toulouse)

In this talk I want to address the themes of the workshop from the perspective of someone who works in lexical semantics. I will argue that phenomena that are ubiquitous in the understanding of language pose problems for a strict Fodor-like separation between conceptual structure related to the non linguistic world and the sort of type system and composition logic needed to handle complex predications that involve coercion or the selection of an aspect of some type of object for predication.

On the time course of semantic/conceptual processing – insights from neutral and emotional words

Christian Dobel (University of Münster)

We found in a series of studies (e.g. Dobel et al., 2010) that the N400 can be taken as a general index for the difficulty of retrieving conceptual knowledge. In contrast, if the sentential context facilitates a visual representation, we also found earlier activity in regions of visual processing arguing for “embodied” semantic representation (Hirschfeld et al., 2010). Similarly, the processing of emotional words is characterized by enhanced activity already around 100 ms (Keuper et al., subm.) which has been often interpreted as attentional effects. It seems, thus, that “meaningful” representations can be achieved by several means with different time course and neural correlates.

Pitfalls in the language-thought distinction: a view on studies of linguistic relativity

Holden Härtl (University of Kassel)

The distinction between linguistic and non-linguistic cognition is a logical prerequisite for the old Whorfian hypothesis that “thought”, i.e. our perception of the world, is structured relative to “language”. In my paper I will critically discuss a selection of current linguistic relativity studies and in how far they indeed access non-linguistic cognition and, in particular, if accusations of a circular reasoning in the interpretation of the observed effects may be justified. Here, the characteristics of the causal relations involved in cognitive differences between linguistic systems need to be carefully considered as well as the philosophically rooted notion of the unavoidability of language.

Is comprehension feasible without validation? Evidence from a Stroop-like paradigm

Maj-Britt Isberner & Tobias Richter (University of Kassel)

Based on the distinction between semantic and world knowledge, comprehension and validation of linguistic information have often been conceptualized as separate stages in a two-step process, with validation being subsequent, optional and strategic. In contrast with this idea, we present evidence from a Stroop-like paradigm suggesting that validation is nonstrategic and fast by showing that readers cannot ignore violations of (easily accessible) world knowledge even if this is in conflict with their processing goal. These results indicate that both access to and evaluation based on world knowledge are obligatory components of language comprehension.

“Accumulating a coin” or “accumulating nuns”: Processing semantic vs. conceptual violations

Barbara Kaup & Claudia Maienborn (University of Tübingen)

The talk is concerned with the distinction between semantic and conceptual knowledge. In a first experiment the final word of a sentence (e.g., sour) was processed more slowly when it brought about a semantic compared to a conceptual violation (These clocks were sour vs. These bananas were sour.). However, similar differences were observed when the respective word pairs (e.g., clocks – sour vs. bananas – sour) were presented in a word priming experiment, suggesting that the results may reflect differences in lexical associations instead of differences in the type of knowledge that is being violated. In our talk we will discuss several alternative ways to empirically investigate the distinction between semantic and conceptual knowledge.


Asher, N. (2011). Lexical meaning in context. A web of words. Cambridge: Cambridge University Press.

Dobel, C., Junghöfer, M., Klauke, B., Breitenstein, C., Pantev, C., Knecht, S. & Zwitserlood, P. (2010). New names for known things: On the association of novel word forms with existing semantic information. Journal of Cognitive Neuroscience, 22, 1251-1261.

Hagoort, P. et al. (2004). Integration of word meaning and world knowledge in Language Comprehension. Science, 304, 438-441.

Hirschfeld, G., Zwitserlood, P. & Dobel, C. (2010). Effects of language comprehension on visual processing - MEG dissociates early perceptual and late N400 effects. Brain and Language, 116, 91-96.

Keuper, K. Zwanzger, P., Zwitserlood, P. & Dobel, C. (subm). How “hate” and “love” differ from “sleep”: Using combined EEG/MEG data and realistic head-modelling to reveal the sources of early cortical responses to emotional words.

Pylkkänen, L., Oliveri, B. & Smart, A.J. (2009). Semantics vs. world knowledge in prefontal cortex. Language and Cognitive Processes, 24, 1313-1334.

Vigliocco, G. & Vinson, D. P. (2007). Semantic representation. In M. G. Gaskell (Ed.), The Oxford Handbook of Psycholinguistics (pp. 195-215). Oxford: Oxford University Press.