Pith. sign in

REVIEW 3 major objections 5 minor 63 references

A Grounded Typology of Word Classes

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Groundedness measures how much of a word's predictability comes from the image it describes, and across 30 languages it ranks word classes from lexical to functional.

desk verdict A genuinely new measurement idea and a valuable 30-language dataset, but the central PMI interpretation rests on an unvalidated model-matching assumption, so the functional-class positive groundedness claim is provisional. read the letter →

arxiv 2412.10369 v1 pith:5Q2D5TJC submitted 2024-12-13 cs.CL cs.CV

Coleman Haley , Sharon Goldwater , Edoardo Ponti This is my paper · ORCID
classification cs.CLcs.CV
keywords groundednesswordclassespartsofspeechlinguistictypologymutualinformationsurprisalmultimodallanguagemodelsimagecaptioning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes 'groundedness', a measure of how much of a word's predictability comes from the image it describes, computed as the difference in surprisal between an image-conditioned captioning model and a matched text-only language model. Applied to word classes in 30 languages, the measure recovers the lexical/functional split: proper nouns, nouns, adjectives, and verbs rank above particles, auxiliaries, conjunctions, determiners, and adpositions, with a nearly total ordering across languages. The paper also finds that functional classes have significantly positive groundedness, which it takes as evidence that grammatical words carry semantic content. This gives typology a quantitative, image-grounded way to study meaning without relying on hand-built semantic categories.

What carries the argument

The carrying machinery is the paired-model surprisal comparison: a multilingual image captioning model $p_\phi(w_t \mid m, w_{<t})$ and a text-only language model $p_\theta(w_t \mid w_{<t})$ trained from the same pretrained weights on the same captions. Groundedness is the negative difference of their surprisals, i.e. the pointwise mutual information with the image, and images act as a language-neutral proxy for meaning. Word-class values come from a Monte Carlo average of token-level groundedness over all tokens carrying that part-of-speech tag, with one-sample permutation tests used to ask whether the class-level mutual information is significantly above zero.

What would settle it

Score the same models on captions paired with mismatched images or with no image; if functional words retain the same positive groundedness in that control, the signal reflects model differences rather than word–image association.

Watch

Extended reading notes

Core claim

The central claim is that groundedness—defined as $\log \frac{p_\phi(w_t \mid m, w_{<t})}{p_\theta(w_t \mid w_{<t})}$ for a word token $w_t$ in context $w_{<t}$ with image $m$—estimates the pointwise mutual information between that word and the image it describes. Averaging this quantity over all tokens tagged with a given part of speech yields an estimate of the mutual information between that word class and images, and across 30 languages the paper finds a consistent, nearly total ranking of word classes: proper nouns, nouns, adjectives, and verbs above particles, auxiliaries, conjunctions, determiners, and adpositions. The paper also reports that functional classes show significantly positive groundedness, which it reads as evidence that grammatical words are not devoid of semantic content.

Load-bearing premise

The entire measure relies on the captioning model and the text-only language model being identical in everything except the image input, so that a larger word probability in the captioning model can only be attributed to information from the image.

Editorial extensions

If this is right

  • Word classes can be treated as a graded lexical-to-functional cline rather than a dichotomy, since estimated marginal means place every part of speech on a continuous groundedness scale.
  • Grammatical words are not semantically inert: determiners, adpositions, auxiliaries, and particles show significant positive groundedness in most of the 30 languages.
  • Images provide a language-neutral meaning representation, so groundedness can be compared across typologically diverse languages directly.
  • The released dataset of per-token and per-class groundedness scores lets other researchers measure groundedness for constructions, morphemes, or finer subclasses without retraining the models.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Using video instead of still images would likely raise verbs' groundedness relative to nouns, since verbs denote temporally extended events; this is a testable consequence of the paper's image-proxy assumption.
  • If groundedness tracks semantic contentfulness, a word undergoing grammaticalization should show falling groundedness as it moves from lexical to functional use, measurable on diachronic corpora.
  • The ranking could be validated behaviorally by testing whether groundedness predicts response times in lexical decision or picture-naming beyond standard concreteness norms.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes 'groundedness,' an empirical measure of a word token's semantic contentfulness defined as the log-probability ratio between an image-conditioned captioning model and a text-only language model (Eq. 1), interpreted as pointwise mutual information between the word and the image. The measure is applied to Universal Dependencies part-of-speech classes across 30 languages using three multilingual image-caption datasets. The authors report that lexical classes are consistently more grounded than functional classes, that the ranking of classes is nearly total (proper nouns, nouns, adjectives, verbs above particles, auxiliaries, conjunctions, determiners, adpositions), and that functional classes show significantly positive groundedness, which they interpret as contradicting the view that functional classes do not convey content. They also report weak-to-moderate correlations with English psycholinguistic concreteness norms and release the groundedness scores.

Significance. If the measure is valid, the paper offers a novel, language-agnostic, corpus-based route to quantifying semantic contentfulness and to testing typological claims about word classes. The strength of the paper is its careful experimental scaffolding: evaluation on three datasets with complementary properties, an explicit attempt to match the training data of the two models, permutation tests with multiple-testing correction, ANOVA variance decomposition, and a public release of the scores. The measure itself is parameter-free: the ranking and class differences are outputs rather than fitted inputs, and the only post-hoc component (the uncertainty-coefficient normalization in Section 5.4) is clearly labeled. However, the central interpretation depends on the two neural models being matched except for the image input, and that premise is not validated with a control condition. The paper's headline claims about functional classes therefore require additional evidence.

major comments (3)
  1. [Section 4, Eq. (1)] The validity of groundedness as an estimate of PMI with the image depends on p_phi and p_theta being matched in every respect except the image input, but this is not the case. The language model is initialized from paligemma-3b-pt-224, a checkpoint that was pretrained on large-scale vision-language data including image tokens, then fine-tuned on COCO-35L captions only. The captioning model uses the same pretrained decoder plus a vision encoder and 256 prepended image tokens. These models therefore differ in architecture, input distribution, and pretraining exposure, not only in the presence of image information. If p_theta systematically underestimates function words in caption-style text, the log-ratio in Eq. (1) will be positive for functional classes even under a null hypothesis of zero image contribution. This directly threatens the Section 5.2 claim that functional classes 'do carry semantic content.' A control condition is needed: for example, computing groundedness with mismatched images (e.g., shuffled image-caption pairs) or comparing against a text-only pretrained LM fine-tuned on the same captions. Without such a check, the PMI interpretation of the reported positive values is not established.
  2. [Section 5.1, Figure 2] The one-sample permutation test randomly flips the signs of observed PMI values and compares the observed mean against the resulting null distribution. This procedure tests the null hypothesis that the PMI distribution is symmetric about zero, not the stated null hypothesis that the mean mutual information is zero. As Figure 3 shows, PMI distributions by part of speech are skewed and heavy-tailed, so a class with a skewed distribution but zero mean could be declared significantly grounded by this test. The authors should use a bootstrap confidence interval for the mean or a one-sample t-test on the average, or explicitly justify why sign-flipping is valid for their null of MI = 0. Since sample sizes are very large, the qualitative conclusions about the ranking may survive, but the reported p-values do not support the stated null as written.
  3. [Section 5.2, Figure 3] The claim of a 'near total ranking' is based on pairwise significance tests over very large token counts, where even tiny mean differences can become significant. The estimated marginal means and pairwise significance do not convey the magnitude or overlap of the class-level distributions. For example, the separation between adpositions and other functional classes, which underlies the claim that adpositions are not semi-lexical, should be reported with effect sizes or confidence intervals for the EMM differences. Without such information, the linguistic interpretation of the ranking (e.g., 'nouns > adjectives > verbs' as a substantive cline) is stronger than the displayed evidence supports.
minor comments (5)
  1. [Introduction] There is a typo in 'pyscho- and neurolinguistics' (should be 'psycho-'), and 'word classs' appears in Section 5.1.
  2. [Section 5.3] The text uses 'ANOV A' instead of 'ANOVA' in the sentence describing the variance decomposition.
  3. [Section 3, Eq. (3)] The expectation notation 'Ep(Ci, m, w<t)' is not defined; clarifying that the average in Eq. (4) is the Monte Carlo estimator would help readers connect the formal definition to the implementation.
  4. [Appendix A.3] The training details state 430,000 steps with a batch size of 4, which is about three epochs over COCO-35L; the relationship between steps, epochs, and dataset size should be made explicit to allow reproducibility.
  5. [Section 5.4] The uncertainty coefficient is introduced as 'the average ratio between LM surprisal and captioning model surprisal,' but the standard definition is a proportional reduction in surprisal; the text should state the formula to avoid ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: groundedness is a direct PMI estimator, and the word-class ranking is an empirical output, not an input.

full rationale

The paper's derivation chain is self-contained rather than circular. Equation 1 defines token groundedness as the log ratio of the captioning-model probability to the language-model probability, which is exactly the pointwise mutual information PMI(w_t; m | w<t); Equation 4 then computes a Monte Carlo average of these PMIs over tokens tagged with each Universal Dependencies class. The lexical-versus-functional ranking and the significance tests in Section 5 are outputs of this estimator applied to three evaluation datasets, not fitted parameters or inputs to the model. The only auxiliary quantity, the uncertainty coefficient in Section 5.4, is explicitly introduced as a normalization of PMI by LM surprisal and is labeled as a secondary analysis, so it does not define the main results. No equation reduces the reported class groundedness to the training objective or to the word-class definitions by construction. The self-citations that appear (e.g., Berger and Ponti 2024; Haley et al. 2024) are background or future-work references and are not load-bearing for the groundedness derivation. The Limitations section raises validity caveats about using images as a meaning proxy and about POS tagging noise, but these are correctness concerns, not circular reductions; they do not alter the verdict that the measure and the reported hierarchy are empirical outputs.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no free parameters; the groundedness measure is a fixed function of two pretrained neural models. The main axioms are the language-agnostic status of images, the conditional independence of POS labels, the comparability of the two models, and the validity of Universal Dependencies tags. None of these are formalized or verified within the paper, so they remain assumptions.

assumptions (4)
  • domain assumption Images are a language-agnostic representation of meaning.
    Section 1 states 'we can use the image as an evidence-based, language-agnostic representation of the shared semantics' underlying sentences across languages.
  • domain assumption The word class label is independent of the meaning representation given the word and its context.
    Section 3 introduces the factorization in Eq. 2: 'We make an assumption that this label is independent of our meaning representation given a word's context.'
  • domain assumption The captioning model and language model are matched in training data and domain, so their log-probability difference estimates PMI.
    Section 4 describes matching the models 'so that they see the same set of captions' and are 'maximally comparable', which is necessary for the surprisal difference to isolate image information.
  • domain assumption Universal Dependencies POS tags are a valid cross-linguistic operationalization of word classes.
    Section 2 states 'we investigate word classes as operationalized in a framework where there is a fixed set of universally applicable word classes, as set out in the Universal Dependencies project'.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Grounded Typology of Word Classes." pith.science (2026). https://pith.science/paper/5Q2D5TJC

@misc{pith2026241210369,
  author       = {Pith},
  title        = {Pith review of: A Grounded Typology of Word Classes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5Q2D5TJC}},
  note         = {Machine review of arXiv:2412.10369}
}
read the original abstract

We propose a grounded approach to meaning in language typology. We treat data from perceptual modalities, such as images, as a language-agnostic representation of meaning. Hence, we can quantify the function--form relationship between images and captions across languages. Inspired by information theory, we define "groundedness", an empirical measure of contextual semantic contentfulness (formulated as a difference in surprisal) which can be computed with multilingual multimodal language models. As a proof of concept, we apply this measure to the typology of word classes. Our measure captures the contentfulness asymmetry between functional (grammatical) and lexical (content) classes across languages, but contradicts the view that functional classes do not convey content. Moreover, we find universal trends in the hierarchy of groundedness (e.g., nouns > adjectives > verbs), and show that our measure partly correlates with psycholinguistic concreteness norms in English. We release a dataset of groundedness scores for 30 languages. Our results suggest that the grounded typology approach can provide quantitative evidence about semantic function in language.

Figures

Figures reproduced from arXiv: 2412.10369 by the authors.

Figure 1
Figure 1. Mean and standard deviation of per-language [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Heatmap of mutual information estimates across parts of speech in thirty languages. Cells show the statistical significance of a word class’s grounded￾ness (MI > 0). Unattested classes are white. Some func￾tional classes display non-significant levels of ground￾edness in several languages, while lexical classes domi￾nantly show highly significant grounding. comprises 30,000 images captioned 5 times each in English, … view at source ↗
Figure 3
Figure 3. Word token level distributions of the groundedness measure (PMI) across all languages and datasets, [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Correlation between human concreteness ratings and type-level groundedness (PMI; left, [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Correlation between English psycholinguistic norms and type-level groundedness (left) or uncertainty [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 31 canonical work pages

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Farrell Ackerman and Robert Malouf. 2013. Morphological Organization : The Low Conditional Entropy Conjecture . Language, 89(3):429--464

    work page 2013

  4. [4]

    Malihe Alikhani and Matthew Stone. 2019. https://doi.org/10.18653/v1/W19-1806 `` caption '' as a coherence relation: Evidence and implications . In Proceedings of the Second Workshop on Shortcomings in Vision and Language, pages 58--67, Minneapolis, Minnesota. Association for Computational Linguistics

    doi:10.18653/v1/w19-1806 2019

  5. [5]

    Backhouse

    A.E. Backhouse. 1984. https://doi.org/10.1016/0024-3841(84)90074-3 Have all the adjectives gone? Lingua, 62(3):169--186

    work page doi:10.1016/0024-3841(84)90074-3 1984

  6. [6]

    Barend Beekhuizen, Julia Watson, and Suzanne Stevenson. 2017. Semantic typology and parallel corpora: S omething about indefinite pronouns. In 39th Annual Conference of the Cognitive Science Society ( C og S ci) , pages 112--117

    work page 2017

  7. [7]

    Yoav Benjamini and Daniel Yekutieli. 2001. https://arxiv.org/abs/2674075 The Control of the False Discovery Rate in Multiple Testing under Dependency . The Annals of Statistics, 29(4):1165--1188

    work page arXiv 2001

  8. [8]

    Uri Berger and Edoardo M. Ponti. 2024. https://arxiv.org/abs/2409.16646 Cross-lingual and cross-cultural variation in image descriptions . Preprint, arXiv:2409.16646

    work page Pith review arXiv 2024

Show all 63 references
  1. [9]

    Lucas Beyer, Andreas Steiner, André Susano Pinto, Alexander Kolesnikov, Xiao Wang, Daniel Salz, Maxim Neumann, Ibrahim Alabdulmohsin, Michael Tschannen, Emanuele Bugliarello, Thomas Unterthiner, Daniel Keysers, Skanda Koppula, Fangyu Liu, Adam Grycner, Alexey Gritsenko, Neil H...

    2024 arXiv

  2. [10]

    Helen Bird, David Howard, and Sue Franklin. 2003. https://doi.org/10.1016/S0911-6044(02)00016-7 Verbs and nouns: The importance of being imageable . Journal of Neurolinguistics, 16(2):113--149

    2003 doi

  3. [11]

    Walter Bisang. 2017. https://doi.org/10.1093/acrefore/9780199384655.013.103 Grammaticalization . In Oxford Research Encyclopedia of Linguistics . Oxford University Press

    2017

  4. [12]

    Geert Booij. 2007. https://doi.org/10.1093/acprof:oso/9780199226245.003.0005 Inflection . In Geert Booij, editor, The Grammar of Words : An Introduction to Linguistic Morphology , pages 99--124. Oxford University Press

    2007

  5. [13]

    Marc Brysbaert, Amy Beth Warriner, and Victor Kuperman. 2014. https://doi.org/10.3758/s13428-013-0403-5 Concreteness ratings for 40 thousand generally known English word lemmas . Behavior Research Methods, 46(3):904--911

    2014 doi

  6. [14]

    Xi Chen, Xiao Wang, Soravit Changpinyo, A. J. Piergiovanni, Piotr Padlewski, Daniel Salz, Sebastian Goodman, Adam Grycner, Basil Mustafa, Lucas Beyer, Alexander Kolesnikov, Joan Puigcerver, Nan Ding, Keran Rong, Hassan Akbari, Gaurav Mishra, Linting Xue, Ashish V. Thapliyal, J...

    2023

  7. [15]

    Christine Chiarello, Connie Shears, and Kevin Lund. 1999. https://doi.org/10.3758/BF03200739 Imageability and distributional typicality measures of nouns and verbs in contemporary English . Behavior Research Methods, Instruments, & Computers, 31(4):603--637

    1999 doi

  8. [16]

    Norbert Corver and Henk Van Riemsdijk. 2001. https://doi.org/10.1515/9783110874006.1 Semi-lexical categories . In Norbert Corver and Henk Van Riemsdijk, editors, Semi-Lexical Categories , pages 1--20. de G ruyter

    2001 doi

  9. [17]

    Ryan Cotterell and Jason Eisner. 2017. https://doi.org/10.18653/v1/P17-1109 Probabilistic typology: Deep generative models of vowel inventories . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1182--119...

    2017 doi

  10. [18]

    Mielke, Jason Eisner, and Brian Roark

    Ryan Cotterell, Sabrina J. Mielke, Jason Eisner, and Brian Roark. 2018. https://doi.org/10.18653/v1/N18-2085 Are all languages equally hard to language-model? In Proceedings of the 2018 Conference of the North A merican Chapter of the Association for Computational Linguistics:...

    2018 doi

  11. [19]

    William Croft. 2002. https://doi.org/10.1017/CBO9780511840579 Typology and Universals , 2nd edition. Cambridge University Press

    2002 doi

  12. [20]

    Manning, Joakim Nivre, and Daniel Zeman

    Marie-Catherine d e Marneffe, Christopher D. Manning, Joakim Nivre, and Daniel Zeman. 2021. https://doi.org/10.1162/coli_a_00402 Universal Dependencies . Computational Linguistics, 47(2):255--308

    2021 doi

  13. [21]

    Catherine Dub \'e , Laura Monetta, Mar \'i a Macarena Mart \'i nez-Cuiti \ n o , and Maximiliano A. Wilson. 2014. https://doi.org/10.7334/psicothema2014.31 Independent effects of imageability and grammatical class in synonym judgement in aphasia. Psicothema, 26(4):449--456

    2014 doi

  14. [22]

    Francis Ferraro, Nasrin Mostafazadeh, Ting-Hao Huang, Lucy Vanderwende, Jacob Devlin, Michel Galley, and Margaret Mitchell. 2015. https://doi.org/10.18653/v1/D15-1021 A survey of current datasets for vision and language research . In Proceedings of the 2015 Conference on Empir...

    2015 doi

  15. [23]

    Simeon Floyd. 2011. https://doi.org/10.1515/lity.2011.003 Re-discovering the Quechua adjective . Linguistic Typology, 15(1):25--63

    2011 doi

  16. [24]

    Spandana Gella, Rico Sennrich, Frank Keller, and Mirella Lapata. 2017. https://doi.org/10.18653/v1/D17-1303 Image pivoting for learning multilingual multimodal representations . In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 28...

    2017 doi

  17. [25]

    Team Gemma. 2024. https://arxiv.org/abs/2403.08295 Gemma: O pen models based on G emini research and technology . Preprint, arXiv:2403.08295

    2024 arXiv

  18. [26]

    Talmy Givon. 1984. Syntax: A Functional-Typological Introduction Vol I. Amsterdam: Benjamins

    1984

  19. [27]

    Joseph Harold Greenberg, editor. 1966. Universals of Language, 2nd edition. Number 37 in The M . I . T . Press Paperback Series. M.I.T Pr, Cambridge, Mass

    1966

  20. [28]

    John Hale. 2001. https://aclanthology.org/N01-1021 A probabilistic E arley parser as a psycholinguistic model . In Second Meeting of the North A merican Chapter of the Association for Computational Linguistics

    2001

  21. [29]

    Ponti, and Sharon Goldwater

    Coleman Haley, Edoardo M. Ponti, and Sharon Goldwater. 2024. https://doi.org/10.15398/jlm.v12i2.351 Corpus-based measures discriminate inflection and derivation cross-linguistically . Journal of Language Modelling, 12(2):477–529

    2024 doi

  22. [30]

    Martin Haspelmath. 2007. https://doi.org/10.1515/LINGTY.2007.011 Pre-established categories don't exist: Consequences for language description and typology . Linguistic Typology , 11(1):119--132

    2007 doi

  23. [31]

    Martin Haspelmath. 2010. https://arxiv.org/abs/40961695 Comparative concepts and descriptive categories in crosslinguistic studies . Language, 86(3):663--687

    2010

  24. [32]

    Martin Haspelmath. 2012. https://doi.org/10.5281/ZENODO.3678496 How to compare major word-classes across the world's languages . UCLA Working Papers in Linguistics, 17:109--130

    2012 doi

  25. [33]

    Henrison Hsieh. 2019. https://doi.org/10.1007/s11049-018-9422-3 Distinguishing nouns and verbs: A Tagalog case study . Natural Language & Linguistic Theory, 37(2):523--569

    2019 doi

  26. [34]

    Daniel Kaufman. 2009. https://doi.org/10.1515/THLI.2009.001 Austronesian Nominalism and its consequences: A Tagalog case study . Theoretical Linguistics , 35(1):1--49

    2009 doi

  27. [35]

    Karin Kipper Schuler, Anna Korhonen, and Susan Brown. 2009. https://aclanthology.org/N09-4007 V erb N et overview, extensions, mappings and applications . In Proceedings of Human Language Technologies: The 2009 Annual Conference of the North A merican Chapter of the Associatio...

    2009

  28. [36]

    Roger Levy. 2008. https://doi.org/10.1016/j.cognition.2007.05.006 Expectation-based syntactic comprehension . Cognition, 106(3):1126--1177

    2008 doi

  29. [37]

    Johan Liljencrants, Bj \"o rn Lindblom, and Bjorn Lindblom. 1972. https://doi.org/10.2307/411991 Numerical Simulation of Vowel Quality Systems : The Role of Perceptual Contrast . Language, 48(4):839--862

    1972 doi

  30. [38]

    Kimberly R Lin, Lisa Wisman Weil, Audrey Thurm, Catherine Lord, and Rhiannon J Luyster. 2022. https://doi.org/10.1177/23969415221085827 Word imageability is associated with expressive vocabulary in children with autism spectrum disorder . Autism & Developmental Language Impairments, 7

    2022 doi

  31. [39]

    Fangyu Liu, Emanuele Bugliarello, Edoardo Maria Ponti, Siva Reddy, Nigel Collier, and Desmond Elliott. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.818 Visually grounded reasoning across languages and cultures . In Proceedings of the 2021 Conference on Empirical Methods i...

    2021 doi

  32. [40]

    Dermot Lynott, Louise Connell, Marc Brysbaert, James Brand, and James Carney. 2020. https://doi.org/10.3758/s13428-019-01316-z The Lancaster Sensorimotor Norms : Multidimensional measures of perceptual and action strength for 40,000 English words . Behavior Research Methods, 5...

    2020 doi

  33. [41]

    Joan Maling and So-Won Kim. 1998. Case assignment in the sipta-construction. In Ross King, editor, Description and Explanation in Korean Linguistics. East Asia Program, Cornell University, Ithaca, NY

    1998

  34. [42]

    Alireza Mohammadshahi, R \'e mi Lebret, and Karl Aberer. 2019. https://doi.org/10.18653/v1/D19-6402 Aligning multilingual word embeddings for cross-modal retrieval task . In Proceedings of the Beyond Vision and LANguage: inTEgrating Real-world kNowledge (LANTERN), pages 11--17...

    2019 doi

  35. [43]

    Byung-Doh Oh and William Schuler. 2024. https://arxiv.org/abs/2406.10851 Leading whitespaces of language models' subword vocabulary poses a confound for calculating word probabilities . Preprint, arXiv:2406.10851

    2024 arXiv

  36. [44]

    Zhang, Coleman Haley, Kenneth Steimel, Han Liu, and Lane Schwartz

    Hyunji Hayley Park, Katherine J. Zhang, Coleman Haley, Kenneth Steimel, Han Liu, and Lane Schwartz. 2021. https://doi.org/10.1162/tacl_a_00365 Morphology matters: A multilingual language modeling analysis . Transactions of the Association for Computational Linguistics, 9:261--276

    2021 doi

  37. [45]

    Andrew K. Pawley. 2006. Where have all the verbs gone? Remarks on the organisation of languages with small, closed verb classes. In 11th Biennial Rice University Linguistics Symposium

    2006

  38. [46]

    Tiago Pimentel and Clara Meister. 2024. https://arxiv.org/abs/2406.14561 How to compute the probability of a word . Preprint, arXiv:2406.14561

    2024 arXiv

  39. [47]

    Tiago Pimentel, Clara Meister, Ethan Wilcox, Kyle Mahowald, and Ryan Cotterell. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.137 Revisiting the optimality of word lengths . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 22...

    2023 doi

  40. [48]

    Frans Plank. 1994. Inflection and derivation. In The Encyclopedia of Language and Linguistics, pages 1671--1679. Elsevier Science and Technology, Amsterdam

    1994

  41. [49]

    Peng Qi, Yuhao Zhang, Yuhui Zhang, Jason Bolton, and Christopher D. Manning. 2020. https://doi.org/10.18653/v1/2020.acl-demos.14 Stanza: A Python Natural Language Processing Toolkit for Many Human Languages . In Proceedings of the 58th Annual Meeting of the Association for Com...

    2020 doi

  42. [50]

    Khapra, Sarath Chandar, and Balaraman Ravindran

    Janarthanan Rajendran, Mitesh M. Khapra, Sarath Chandar, and Balaraman Ravindran. 2016. https://doi.org/10.18653/v1/N16-1021 Bridge correlational neural networks for multilingual multimodal representation learning . In Proceedings of the 2016 Conference of the North A merican ...

    2016 doi

  43. [51]

    Alexander Rauhut. 2023. https://refubium.fu-berlin.de/handle/fub188/40321?show=full Quantitative Aspects of the Word Class Continuum in English . Ph. D . thesis, Freie Universität Berlin

    2023

  44. [52]

    Norvin Richards. 2009. https://doi.org/10.1515/THLI.2009.008 Nouns, verbs, and hidden structure in Tagalog . Theoretical Linguistics, 35(1):139--152

    2009 doi

  45. [53]

    Eva Schultze-Berndt . 2000. Simple and Complex Verbs in Jaminjung : A Study of Event Categorisation in an Australian Language . Ph.D. thesis, Radboud University, Nijmegen

    2000

  46. [54]

    Scott, Anne Keitel, Marc Becirspahic, Bo Yao, and Sara C

    Graham G. Scott, Anne Keitel, Marc Becirspahic, Bo Yao, and Sara C. Sereno. 2019. https://doi.org/10.3758/s13428-018-1099-3 The Glasgow Norms : Ratings of 5,500 words on nine scales . Behavior Research Methods, 51(3):1258--1270

    2019 doi

  47. [55]

    Smith and Roger Levy

    Nathaniel J. Smith and Roger Levy. 2013. https://doi.org/10.1016/j.cognition.2013.02.013 The effect of word predictability on reading time is logarithmic . Cognition, 128(3):302--319

    2013 doi

  48. [56]

    Forthcoming

    Adrian Staub. Forthcoming. https://doi.org/10.1146/annurev-linguistics-011724-121517 Predictability in Language Comprehension : Prospects and Problems for Surprisal . Annual Review of Linguistics

    doi

  49. [57]

    Thapliyal, Jordi Pont Tuset, Xi Chen, and Radu Soricut

    Ashish V. Thapliyal, Jordi Pont Tuset, Xi Chen, and Radu Soricut. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.45 Crossmodal-3600: A Massively Multilingual Multimodal Evaluation Dataset . In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Proce...

    2022 doi

  50. [58]

    Henri Theil. 1970. https://arxiv.org/abs/2775440 On the Estimation of Relationships Involving Qualitative Variables . American Journal of Sociology, 76(1):103--154

    1970

  51. [59]

    David John Weber. 1983. A Grammar of Huallaga (Huanuco) Quechua . Ph.D. thesis, University of California, Los Angeles, United States -- California

    1983

  52. [60]

    Wilcox, Tiago Pimentel, Clara Meister, Ryan Cotterell, and Roger P

    Ethan G. Wilcox, Tiago Pimentel, Clara Meister, Ryan Cotterell, and Roger P. Levy. 2023. https://doi.org/10.1162/tacl_a_00612 Testing the predictions of surprisal theory in 11 languages . Transactions of the Association for Computational Linguistics, 11:1451--1470

    2023 doi

  53. [61]

    Tianxing Wu, Chaoyu Gao, Lin Li, and Yuxiang Wang. 2022. https://doi.org/10.3390/app121910107 Leveraging multi-modal information for cross-lingual entity matching across knowledge graphs . Applied Sciences, 12(19)

    2022 doi

  54. [62]

    Hwang, Amy X

    Andre Ye, Sebastin Santy, Jena D. Hwang, Amy X. Zhang, and Ranjay Krishna. 2024. https://arxiv.org/abs/2310.14356 Computer vision datasets and models exhibit cultural and linguistic diversity in perception . Preprint, arXiv:2310.14356

    2024 arXiv

  55. [63]

    X. Zhai, B. Mustafa, A. Kolesnikov, and L. Beyer. 2023. https://doi.org/10.1109/ICCV51070.2023.01100 Sigmoid loss for language image pre-training . In 2023 IEEE / CVF International Conference on Computer Vision ( ICCV ) , pages 11941--11952, Los Alamitos, CA, USA. IEEE Compute...

    2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.