https://sites.google.com/view/cogalex-viii-2024/home
30th International Conference on Computational Linguistics
Torino, Italy, May 20, 2024
organizers: Zock, M., Chersoni, E., Yu-Yin Hsu, Y-Y, & S. De Deyne
"In the beginning was the word" (Book of Genesis 1:1)
"Without grammar very little can be conveyed, without vocabulary, nothing can be conveyed."
(Wilkins, 1972)
Landscape, Context, and, Problems
No doubt, words matter, yet the way we look at them and the place where they are stored (lexicon) has changed dramatically over the last few decades. While in the past considered only as an appendix to grammar, the lexicon has now moved to center stage, which makes sense. Indeed, there is hardly any language-related task that can be carried out without it.
As to be expected, different views have emerged. Construction grammarians (Croft, 2001; Langacker, 2005). challenge the traditional grammar and lexicon divide, viewing the two as a continuum (Diessel, 2019; Hoffmann, & Trousdale, 2013, Taylor, 2012). Lexicographers working in the Meaning-Text framework integrate grammatical and semantic information in the lexicon (Melçuk, 2006). Psychologists take a completely different stance. Rather than considering words or dictionaries as static entities (database view), they view them as a multi-layered dynamic network, akin to the human brain, whose 'nodes' and 'links' may change their weights (connection strengths) over time.
Regardless of the theoretical stance, many of us are eager to find out how words are acquired, represented, and organized in books (dictionaries), computers, or the human brain (Zock, 2015).. The more practically minded scholar may also wonder what a lexical resource (dictionaries, thesauri, or ontologies) should look like, and how it should be built to support reading/writing, or the higher mental processes following or preceding them, thinking (understanding, conceptualizing, brainstorming, and message planning).
Most laymen and linguists seem to hold the view that words are products, i.e., holistic entities. This view is fine for practical purposes and offline processing like searching in a dictionary, or navigation in a lexical resource. However, this view is inadequate if we deal with online processing (speaking or writing). Word access, or word production by humans, is a process, whose final products, words, have been synthesized over time (Indefrey & Levelt, 2004). Starting from meanings, the speaker activates lemmata, abstract lexical forms devoid of phonological information, to 'retrieve' only then phonological forms: sounds, syllables, and phonemes. Like all processes, word production takes time (around 300 milliseconds per word). It is done stepwise (Dell, 1986, Levelt et al., 1999), and there is no guarantee that its output will be perfect, speech errors (Fromkin, 1980), delays of various sorts including silence, or tip-of-the-tongue problems (Brown, 2012) are evidence of the contrary.
"Speech is normally produced at a rate of about two to three words per second" (Levelt, 1989). This is quite an achievement, given that the average 20-year-old knows about 42,000 lemmata (Brysbaert et al., 2018). Indeed, the speed at which our brain can 'locate' a specific word within such a huge store (the entire lexicon) is intriguing. This is one of the reasons why so many people are interested in the mental lexicon (Aitchison, 2003). We would like to understand its structure and functioning, and we may wonder whether the brain or certain aspects of its functioning could be used as a blueprint for the dictionary of tomorrow (Zock et al., 2022). Yet, strange as it may be, the followers of those having contributed most to the understanding of the mental lexicon (Levelt, et al., 1989; Miller et al., 1988; Dell, 1986), don't seem to communicate much with each other, and one may wonder why.
The reason may be the following. The respective research of these pioneers is based on two different, yet complementary viewpoints and assumptions. One starts from concepts, and asks questions about the mental processes that determine how the corresponding linguistic forms are gradually synthesized (Levelt et al., 1999; Dell, 1986), while the other starts from word forms and is mostly concerned about how words represent meaning as part of complex relational knowledge system (Miller et al., 1988; Fellbaum, 1998). Given some input, the resource is meant to help the user find among the direct neighbors (associated words) the one he is looking for (synonym, antonym, hypernym, ...).
While the first approach consists of activating the relevant nodes in a multilayered network, simulating the normal course of action when trying to produce a word (automatic processing), the second consists of navigating in a lexical network (topological view), which corresponds more to offline processing, i.e., deliberate search in a lexical resource (Zock et al., 2010). What matters most to the members of this latter group are the answers to the following questions: given some input, what are its direct neighbors, and how do they relate to the source word (input)? Put differently, this community is trying to build a map of the mental lexicon (lexical graphs or association networks) to support word-finding.
While both communities think in terms of networks, they both make quite different assumptions concerning the reality of words. According to the representatives of the first group, they are decomposed (meaning, form, sound) requiring recomposition (synthesis), while for the second (relational view) words exist right from the start as holistic entities (dictionary view), though linked in various ways: topically (Roget, 1852), semantically (Fellbaum, 1998) or other sorts of associations (Kiss, 1968; Schvaneveldt, 1989; Nelson, McEvoy, & Schreiber, 1999). It is this linkage that sets this group apart from traditional, alphabetically organized lexicons.
It seems that the two communities work on different planes (vertical/horizontal) and on different time scales, one being fast, and the other one slow (Kahneman, 2011). Psychologists describe the way words are synthesized in real-time (online processing), while computational linguists try to present a map of the (collective) ‘mental lexicon’, allowing for offline processing (navigation).
Working on an extremely small scale (generally fewer than 100 words), psychologists studying the time course of word production (Kerr et al., 2023) cannot offer us a usable resource (map) all the more as this is not their goal, while lexicographers still have a hard time dealing with the diversity of inputs at the onset of the search and the variety of navigational strategies. To overcome these problems, they would need not only to spend more time observing real users (Lew, 2001, de Schryver & Joffe, 2004), but also try to create a much larger set of associations than those available in WordNet (Fellbaum, 1998). To this end, we need to draw on a large variety (well-balanced set) of corpora. A big step in this direction has been made by the creators of BABELNET (Navigli, et al., 2021) and JeuxDeMots (Lafourcade & Joubert, 2015). Actually, the list of lexical graphs and word association thesauri keeps growing (Zock & Biemann, 2020). By and large, the availability of big data and sophisticated corpus processing tools (Kilgarriff et al., 2008) allows us nowadays to create new resources and study all kinds of interesting lexicon-related problems (Zhang, 2023, Grefenstette, 2008).
More recently a new set of techniques has taken the world by storm: Large Language Models (LLMs), Chat-GPT (Riedl, 2023) being the best known. Causing a gold rush and radical paradigm shift, LLMs had an impact in many areas, including AI, NLP, and beyond, even the lexicon (de Schryver, 2023, Lew, 2023). Unlike distributional semantics (Lenci & Sahlgren, 2023), LLMs focus not only on words but also on units of different sizes: sublexical elements, flemmas (word families). Since all of them seem to be relevant to explaining the functioning of language we should think twice next time about what to focus on lemmas, words, or flemmas (see Stoeckel et al., 2020).
As one can see, several communities are concerned with the cognitive aspects of the lexicon, theoretical and applied linguists (lexicographers), computer scientists (corpus linguistics), psychologists (association networks), and specialists working on complex graphs (Siew, et al., 2019; De Deyne et al., 2016; Wilks & Meara, 2002). They all could make precious contributions while benefiting from each other’s work. Alas, this is still not yet quite the case, which is why we keep organizing this kind of workshop.
On a more positive note, in the recent past, there have many new, interesting developments. They have huge potential, increasing tremendously the benefits of learning from each other and working together. Yet, it’s up to us to take our chances, to reach out, and to cross the borders of our discipline. So, what are we waiting for?
Aitchison, J. (2003). Words in the Mind: an Introduction to the Mental Lexicon. Oxford: Blackwell.
Brown, A. S. (2012). The tip of the tongue state. Taylor & Francis.
Brysbaert M., Stevens M., Mandera P. & Keuleers E. (2016). How many words do we know? Practical estimates of vocabulary size dependent on word definition, the degree of language input and the participant’s age. Front. Psychol. 7:1116. doi: 10.3389/fpsyg.2016.01116
Croft, W. (2001). Radical construction grammar: Syntactic theory in typological perspective. Oxford University Press, USA.
De Deyne, S., Verheyen, S. & Storms, G. (2016). Structure and organization of the mental lexicon: A network approach derived from syntactic dependency relations and word associations. In Mehler, A., et al. (Eds.). Towards a theoretical framework for analyzing complex linguistic networks (pp. 47–79). Berlin: Springer.
Dell G. S. (1986). A spreading activation theory of retrieval in language production. Psychological Review, 93:283–321.
De Schryver, G. M. (2023). Generative AI and Lexicography: The Current State of the Art Using ChatGPT. International Journal of Lexicography, 36(4), 355-387.
De Schryver, G-M. & D. Joffe. (2004). On how electronic dictionaries are really used. Williams, G. & S. Vessier (Eds.). Proceedings of the 11th EURALEX Congress, Lorient, France, 187-196.
Diessel, H. (2019). The grammar network. How Linguistic Structure is Shaped by Language Use. Cambridge University Press.
Fellbaum, C. (Ed.) (1998). WordNet: An electronic lexical database and some of its applications. Cambridge: MIT Press.
Fromkin V. (ed.). (1980). Errors in linguistic performance: Slips of the tongue, ear, pen, and hand. San Francisco: Academic Press.
Grefenstette, G. (2008). The Future of Linguistics and Lexicographers: Will there be Lexicographers in the Year 3000? In Fontenelle, T. (Ed.). Practical lexicography: a reader. OUP Oxford, 307-324
Hoffmann, T. & Trousdale, G. (Eds.), (2013). The Oxford handbook of construction grammar. Oxford University Press.
Indefrey, P. & Levelt, W. J. (2004). The spatial and temporal signatures of word production components. Cognition, 92(1-2), 101-144.
Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus & Giroux, New York
Kerr, E., Ivanova, B. & Strijkers, K. (2023). Lexical access in speech production: Psycho-and neurolinguistic perspectives on the spatiotemporal dynamics. In Hartsuiker, R. J. & Strijkers, K. (Eds.). Language Production. Routledge.(pp. 32-65). Routledge.
Kilgarriff, A., Rychly, P., Smrž, P. & Tugwell, D. (2008). The sketch engine. In Fontenelle, T. (Ed.). (2008). Practical lexicography: a reader. OUP Oxford, 297-306.
Kiss, G. R. (1968). Words, associations, and networks. Journal of Verbal Learning and Verbal Behavior, 7(4), 707-713.
Lafourcade, M. & Joubert, A. (2015). TOTAKI: a help for lexical access on the TOT problem. In Gala, N., Rapp, R. & Bel-Enguix, G. (Eds.). Language production, cognition, and the lexicon. Springer. 95-112.
Langacker, R. W. (2005). Construction grammars: Cognitive, radical, and less so. In T. Hoffmann & G. Trousdale (Eds.), Handbook of Construction Grammar (pp. 31-71). Oxford: Oxford University Press.
Lenci, A., & Sahlgren, M. (2023). Distributional semantics. Cambridge University Press.
Levelt W. (1989). Speaking: From Intention to Articulation. MIT Press, Cambridge, MA
Levelt W., Roelofs A. & Meyer, A. (1999). A theory of lexical access in speech production. Behavioral and Brain Sciences, 22, 1-75.
Lew, R. (2023). ChatGPT as a COBUILD lexicographer. Humanities and Social Sciences Communications, 10(1), 1-10.
Lew, R. (2011). Studies in dictionary use: Recent developments. International Journal of Lexicography, 24(1), 1-4.
Meara, P. (2009). Connected words: Word associations and second language vocabulary acquisition (Vol. 24). John Benjamins Publishing.
Mel’čuk, I. (2006). Explanatory combinatorial dictionary. In, Giandomenico SICA (ed.), 2006, Open problems in Linguistic and lexicography, Monza (Italy): Polimetrica, pp. 225-355.
Miller, G., Fellbaum, C., Kegl, J. & Miller, K. (1988). WordNet: An Electronic Lexical Reference System Based on Theories of Lexical Memory. Revue québécoise de linguistique, vol. 17, n° 2, pp. 181-212.
Navigli, R., Bevilacqua, M., Conia, S., Montagnini, D., & Cecconi, F. (2021). Ten years of BabelNet: A survey. In IJCAI (International Joint Conferences on Artificial Intelligence Organization). 4559-4567. https://doi.org/10.24963/ijcai.2021/620
Nelson, D. L., McEvoy, C. L. & Schreiber, T. A. (1999). The University of South Florida Word Association, Rhyme and Fragment Norms. (http://w3.usf.edu/FreeAssociation/Intro.html)
Riedl, M. (2023). A very gentle introduction to large language models without the hype. Medium
Roget, P. (1852) Thesaurus of English Words and Phrases, Longman, London
Schvaneveldt, R. editor. (1989). Pathfinder Associative Networks: Studies in knowledge organization. Norwood. N.J.
Siew, C. S., Wulff, D. U., Beckage, N. M. & Kenett, Y. N. (2019). Cognitive Network Science: A review of research on cognition through the lens of network representations, processes, and dynamics. (https://www.hindawi.com/journals/complexity/2019/2108423/)
Stoeckel, T., Ishii, T. & Bennett, P. (2020). Is the lemma more appropriate than the flemma as a word-counting unit? Applied Linguistics, 41(4), 601-606. https://doi.org/10.1093/applin/amy059
Taylor, J. R. (2012), The Mental Corpus: How Language is Represented in the Mind, Oxford: Oxford University Press.
Wilkins, D. (1972) Linguistics in Language Teaching. Edward Arnold.
Wilks, C. & Meara, P. (2002). Untangling word webs: Graph theory and the notion of density in second language word association networks. Second Language Research, 18(4), 303-324.
Zhang, Y. (2023). Computational Lexicography: From the Electronic to Digitalized Age. In Sin-Wai, C. (Eds.). Routledge encyclopedia of translation technology. Taylor & Francis Group. (pp. 483-497).
Zock, M., de Deyne, S., Stella, M., & Pirrelli, V. (2022). The Mental Lexicon, Blueprint of the Dictionaries of Tomorrow: Cognitive Aspects of the Lexicon. Frontiers in Artificial Intelligence, 5, 945705.
Zock, M. & Biemann, C. (2020). Comparison of different lexical resources with respect to the tip-of-the-tongue problem. Journal of Cognitive Science, 21(2), 193-252.
Zock, M. (2015). Words in Books, Computers, and the Human Mind. Introduction to 'Special Issue' of Journal of Cognitive Science. 16-4: 355-378, DOI : 10.17791/jcs.2015.16.4.355
Zock, M., Ferret, O. & Schwab, D. (2010). Deliberate word access: an intuition, a roadmap and some preliminary empirical results. International Journal of Speech Technology, 13, 201-218.