Hostname: page-component-586b7cd67f-dlnhk Total loading time: 0 Render date: 2024-11-23T01:57:01.043Z Has data issue: false hasContentIssue false

Natural logic and baby LoTH

Published online by Cambridge University Press:  28 September 2023

Irene Canudas-Grabolosa
Affiliation:
Department of Psychology, Department of Linguistics, Harvard University, Cambridge, MA, USA [email protected]
Ana Martín-Salguero
Affiliation:
Center for Brain and Cognition, Universitat Pompeu Fabra, Barcelona, Spain [email protected] Cognitive Neuroimaging Unit, CEA, INSERM, Université Paris-Saclay, NeuroSpin Center, Gif/Yvette, France
Luca L. Bonatti
Affiliation:
Center for Brain and Cognition, Universitat Pompeu Fabra, Barcelona, Spain [email protected] ICREA, Barcelona, Spain [email protected]

Abstract

Language-of-thought hypothesis (LoTH) is having a profound impact on cognition studies. However, much remains unknown about its basic primitives and generative operations. Infant studies are fundamental, but methodologically very challenging. By distilling potential primitives from work in natural-language semantics, an approach beyond the corset of standard formal logic may be undertaken. Still, the road ahead is challenging and long.

Type
Open Peer Commentary
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press

Fodor had the gift of conceiving extremely simple ideas with extremely deep and rich consequences. Language-of-thought hypothesis (LoTH) is perhaps the best, but not the unique, example of this gift. Quilty-Dunn et al.'s article is a very forceful testimony of how lively and far-reaching LoTH is. The very fact that they use a cluster of properties that prescinds from most traditional arguments for LoT is in itself a proof of its richness. At the same time, as it is clear in the target article, like other cases (modularity witness it), Fodor's LoTH was more a research program than an hypothesis; in his words, it's probably a genus, but, we would add, one whose actual species are still barely known. This dearth of knowledge is particularly acute for one of the fundamental issues in characterizing LoT(s): Identify the basic primitives available endogenously in human thinking. In adults, a recent work investigated modular LoTs defined over various domains, proposing primitives and compositional routines (Al Roumi, Marti, Wang, Amalric, & Dehaene, Reference Al Roumi, Marti, Wang, Amalric and Dehaene2021; Dehaene, Al Roumi, Lakretz, Planton, & Sablé-Meyer, Reference Dehaene, Al Roumi, Lakretz, Planton and Sablé-Meyer2022; Planton et al., Reference Planton, van Kerkoerle, Abbih, Maheu, Meyniel, Sigman and Dehaene2021; Sablé-Meyer et al., Reference Sablé-Meyer, Fagot, Caparos, van Kerkoerle, Amalric and Dehaene2021; Sablé-Meyer, Ellis, Tenenbaum, & Dehaene, Reference Sablé-Meyer, Ellis, Tenenbaum and Dehaene2022). However exciting and important to characterize human singularity, these theories do not clarify the origins of LoT or its role in general human cognition. They can check out all the list of properties in Quilty-Dunn et al.'s cluster, and yet remain confined to the specific domain they have been tested, in adults. They are compatible with the fact that language interactions, or instruction, contributes to their appearance.

Although there is little doubt that when linguistic competence kicks in, human language competence is explained by reference to a system of structures encompassing many properties of a general LoT, the crucial open questions are whether properties of general thinking are somehow imported from linguistic structures, as many would hold (Carruthers, Reference Carruthers2002; Spelke, Reference Spelke, Gentner and Goldin-Meadow2003), or else are inherent properties of the mind, and if they encompass the logical concepts that make LoT cross-domain and compositional. Progress on these questions can be achieved by investigating the existence and nature of the logical primitives available to preverbal infants, who are likely not affected by instructions or massive language experience. Unfortunately, these investigations can be counted on the fingers of one hand. They are also very difficult, as they require creating scenes deprived of verbal cues that likely embed logical inferences, something that at best can be supported by arguments to the best explanation. We thank Quilty-Dunn et al., who agree with us that a baby-LoTH has the upper hand relative to alternative theories, but they are more optimistic than us: Alternative explanations, perhaps compatible, perhaps incompatible with some declination of LoTH (Leahy & Carey, Reference Leahy and Carey2020), exist and have to be addressed experimentally. Furthermore, an unified explanation of the early putative indications of logical thinking (Cesana-Arlotti et al., Reference Cesana-Arlotti, Martín, Téglás, Vorobyova, Cetnarski and Bonatti2018; Cesana-Arlotti, Kovács, & Téglás, Reference Cesana-Arlotti, Kovács and Téglás2020; Cesana-Arlotti, Téglás, & Bonatti, Reference Cesana-Arlotti, Téglás and Bonatti2012; Cesana-Arlotti, Varga, & Téglás, Reference Cesana-Arlotti, Varga and Téglás2022) and the later failures at making action plans consistent with it (Feiman, Mody, & Carey, Reference Feiman, Mody and Carey2022; Leahy, Huemer, Steele, Alderete, & Carey, Reference Leahy, Huemer, Steele, Alderete and Carey2022; Mody & Carey, Reference Mody and Carey2016) is still missing. All these issues require painstaking research.

As baby LoTH supporters, we believe that the most serious question remains the identification of a plausible repertoire of early LoT primitives. Short of the success in disjunctive reasoning (Cesana-Arlotti et al., Reference Cesana-Arlotti, Martín, Téglás, Vorobyova, Cetnarski and Bonatti2018), little exists about other logical components of an LoT, while some arguably plausible candidates – for example, simple relations such as “Same/Different” – do not seem to be supported (Hochmann, Reference Hochmann2022; Hochmann et al., Reference Hochmann, Tuerk, Sanborn, Zhu, Long, Dempster and Carey2017; Hochmann, Mody, & Carey, Reference Hochmann, Mody and Carey2016). Where to look for plausible candidates? The naive approach we, our collaborators, and others have taken has been to follow the guidance of formal logic, hence looking for connectives, quantifiers, or Boolean concepts. This is a plausible approach, but deep down is based on the arbitrary assumption that the forms of human thoughts comply with descriptions largely developed for mathematical elegance rather than for psychological reality. Whether thought fits the well-defined but rigid corset of logical systems is far from granted. Indeed, even with the baby LoT case discussed by Quilty-Dunn et al., which we contributed to develop (Cesana-Arlotti et al., Reference Cesana-Arlotti, Martín, Téglás, Vorobyova, Cetnarski and Bonatti2018), it is still not clear whether infants’ mental representations involve disjunctions (A or B), possible alternatives (maybe A, maybe B), or quantified representations (“unknown x”).

Another possible approach is to invert the relation between natural languages and thought. Rather than regarding them as the origin of logical abilities in thought, one could look at their semantics as crystalized repositories of thought primitives. Under this perspective, LoT primitives may well differ from those familiar from logic. For example, all studied languages contain polarity items, but these have no place in logical systems. Yet, undoubtedly, their behavior is “logical”; perhaps they signal the presence of primitive logical operations that are at the source of their widespread presence in natural languages. Likewise, operations such as exhaustification, which seems to be necessary to explain many patterns of implications in language (Chierchia, Reference Chierchia2013), do not exist in logic, but could potentially be present in an LoT, as a primitive operation defined over sets and set relations (another area which, with few exceptions [Feigenson & Halberda, Reference Feigenson and Halberda2004, Reference Feigenson and Halberda2008; Zosh, Halberda, & Feigenson, Reference Zosh, Halberda and Feigenson2011], is still poorly known). Tense and modal structure can also offer case studies for the identification of potential logical primitives.

We feel that the interaction between natural-language semantics and psychology can be a fruitful way to unlock basic potential primitives of thought. From there on, painful and long case studies have to be developed to trace back their nonverbal origins in infant LoT(s). It took about 50 years to transform LoTH into a fruitful research program; we bet that it won't take much less to go from the acceptance of the genus LoT to the discovery of its species and their common origins in preverbal infants. Nonetheless, we feel, the potential for new insights and discoveries makes this endeavor worthwhile undertaking.

Financial support

This work was supported by the Ministerio de Ciencia e Innovación Grant PID2019-108494GB-I00 to L. L. B.

Competing interest

None.

References

Al Roumi, F., Marti, S., Wang, L., Amalric, M., & Dehaene, S. (2021). Mental compression of spatial sequences in human working memory using numerical and geometrical primitives. Neuron, 109(16), 26272639, e4.CrossRefGoogle ScholarPubMed
Carruthers, P. (2002). The cognitive functions of language. Behavioral and Brain Sciences, 25(6), 657726.CrossRefGoogle ScholarPubMed
Cesana-Arlotti, N., Kovács, Á. M., & Téglás, E. (2020). Infants recruit logic to learn about the social world. Nature Communications, 11(1), 19.CrossRefGoogle ScholarPubMed
Cesana-Arlotti, N., Martín, A., Téglás, E., Vorobyova, L., Cetnarski, R., & Bonatti, L. L. (2018). Precursors of logical reasoning in preverbal human infants. Science (New York, N.Y.), 359(6381), 12631266.CrossRefGoogle ScholarPubMed
Cesana-Arlotti, N., Téglás, E., & Bonatti, L. L. (2012). The probable and the possible at 12 months: Intuitive reasoning about the uncertain future. Advances in Child Development and Behavior, 43, 125.CrossRefGoogle ScholarPubMed
Cesana-Arlotti, N., Varga, B., & Téglás, E. (2022). The pupillometry of the possible: An investigation of infants' representation of alternative possibilities. Philosophical Transactions of the Royal Society of London, Series B: Biological Sciences, 377(1866), 20210343.CrossRefGoogle ScholarPubMed
Chierchia, G. (2013). Logic in grammar: Polarity, free choice, and intervention. MIT Press.CrossRefGoogle Scholar
Dehaene, S., Al Roumi, F., Lakretz, Y., Planton, S., & Sablé-Meyer, M. (2022). Symbols and mental programs: A hypothesis about human singularity. Trends in Cognitive Science, 26(9), 751766.CrossRefGoogle ScholarPubMed
Feigenson, L., & Halberda, J. (2004). Infants chunk object arrays into sets of individuals. Cognition, 91(2), 173190.CrossRefGoogle ScholarPubMed
Feigenson, L., & Halberda, J. (2008). Conceptual knowledge increases infants' memory capacity. Proceedings of the National Academy of Sciences of the United States of America, 105(29), 99269930.CrossRefGoogle ScholarPubMed
Feiman, R., Mody, S., & Carey, S. (2022). The development of reasoning by exclusion in infancy. Cognitive Psychology, 135, 101473.CrossRefGoogle ScholarPubMed
Hochmann, J. R. (2022). Representations of abstract relations in infancy. Open Mind, 6, 291310.CrossRefGoogle ScholarPubMed
Hochmann, J. R., Mody, S., & Carey, S. (2016). Infants' representations of same and different in match- and non-match-to-sample. Cognitive Psychology, 86, 87111.CrossRefGoogle ScholarPubMed
Hochmann, J.-R., Tuerk, A. S., Sanborn, S., Zhu, R., Long, R., Dempster, M., & Carey, S. (2017). Children’s representation of abstract relations in relational/array match-to-sample tasks. Cognitive Psychology, 99, 1743.CrossRefGoogle ScholarPubMed
Leahy, B., Huemer, M., Steele, M., Alderete, S., & Carey, S. (2022). Minimal representations of possibility at age 3. Proceedings of the National Academy of Sciences of the United States of America, 119(52), e2207499119.CrossRefGoogle ScholarPubMed
Leahy, B. P., & Carey, S. E. (2020). The acquisition of modal concepts. Trends in Cognitive Sciences, 24(1), 6578.CrossRefGoogle ScholarPubMed
Mody, S., & Carey, S. (2016). The emergence of reasoning by the disjunctive syllogism in early childhood. Cognition, 154, 4048.CrossRefGoogle ScholarPubMed
Planton, S., van Kerkoerle, T., Abbih, L., Maheu, M., Meyniel, F., Sigman, M., … Dehaene, S. (2021). A theory of memory for binary sequences: Evidence for a mental compression algorithm in humans. PLoS Computational Biology, 17(1), e1008598.CrossRefGoogle ScholarPubMed
Sablé-Meyer, M., Ellis, K., Tenenbaum, J., & Dehaene, S. (2022). A language of thought for the mental representation of geometric shapes. Cognitive Psychology, 139, 101527.CrossRefGoogle ScholarPubMed
Sablé-Meyer, M., Fagot, J., Caparos, S., van Kerkoerle, T., Amalric, M., & Dehaene, S. (2021). Sensitivity to geometric shape regularity in humans and baboons: A putative signature of human singularity. Proceedings of the National Academy of Sciences of the United States of America, 118(16), e2023123118.CrossRefGoogle ScholarPubMed
Spelke, E. S. (2003). What makes us smart? Core knowledge and natural language. In Gentner, D. & Goldin-Meadow, S. (Eds.), Language in mind: Advances in the study of language and thought (pp. 277311). MIT Press.Google Scholar
Zosh, J. M., Halberda, J., & Feigenson, L. (2011). Memory for multiple visual ensembles in infancy. Journal of Experimental Psychology: General, 140(2), 141158.CrossRefGoogle ScholarPubMed