pith. sign in

arxiv: 2606.27206 · v1 · pith:HD2IMWG7new · submitted 2026-06-25 · 💻 cs.CL

Syntactic Belief Update as the Driver of Garden Path Processing Difficulty

Alan Zhou , Milo\v{s} Stanojevi\'c , John T. Hale This is my paper

Pith reviewed 2026-06-26 04:35 UTC · model grok-4.3

classification 💻 cs.CL
keywords garden path sentencessyntactic belief updateRényi divergencesentence processingreading timessurprisalpsycholinguisticssyntactic parsing
0
0 comments X

The pith

Syntactic belief updates measured by Rényi divergence on tree distributions explain garden path reading difficulty better than lexical surprisal.

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

The paper argues that sentence processors maintain a probability distribution over possible syntactic trees. A garden path forces a large revision of this distribution at the critical word, and the size of the revision is quantified by generalized Rényi divergence. This divergence depends on the identity of the words but is independent of their lexical probabilities, unlike standard surprisal. The resulting syntactic belief update measure fits observed human reading times on garden path sentences more closely than lexical surprisal does. The work points toward non-lexical alternatives for modeling processing difficulty.

Core claim

The paper claims that the magnitude of the update to a probability distribution over syntactic trees, measured by generalized Rényi divergence after each word, accounts for the processing difficulty observed in garden path sentences and yields a better fit to human reading time data than lexical surprisal.

What carries the argument

Syntactic belief update: the change in a probability distribution over syntactic trees after each new word, quantified by generalized Rényi divergence between the pre-update and post-update distributions.

If this is right

  • The syntactic belief update predicts larger reading time increases precisely at the disambiguating word in garden path sentences.
  • The update measure remains sensitive to lexical item identity while staying fully independent of lexical item probabilities.
  • This approach supplies an alternative predictor for cases where lexical surprisal underpredicts difficulty.
  • It motivates examining purely syntactic or structural sources of processing cost separate from word-level probabilities.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the divergence measure is correct, incremental parsers that track multiple syntactic hypotheses should show measurable cost when probability mass shifts between them.
  • The independence from lexical probabilities allows experiments that hold word frequencies constant while varying syntactic ambiguity to isolate the effect.
  • The same update mechanism could be tested on other reanalysis phenomena, such as temporary ambiguities in relative clauses or coordination.
  • Computational models of sentence processing could implement this divergence as an explicit cost signal without requiring separate lexical surprisal terms.

Load-bearing premise

A probability distribution over syntactic trees can be maintained and updated after each word independently of lexical probabilities yet still sensitive to word identity, and the size of the Rényi divergence directly indexes processing difficulty.

What would settle it

An experiment that measures reading times on garden path sentences and finds no reliable correlation with the magnitude of the generalized Rényi divergence on syntactic tree distributions after controlling for lexical factors.

Figures

Figures reproduced from arXiv: 2606.27206 by Alan Zhou, John T. Hale, Milo\v{s} Stanojevi\'c.

Figure 1
Figure 1. Figure 1: Empirical garden path effects– defined as the total slowdown over the critical region between a garden [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Sampled correlations between predicted slowdowns and and empirical slowdowns estimated from using a [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of empirical and predicted garden path effects from our alternative baselines, fit on critical [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Sampled correlations between predicted slowdowns and and empirical slowdowns from our alternative [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Frequentist comparison of Rényi Divergence and Rényi Cross-Entropy for different values of [PITH_FULL_IMAGE:figures/full_fig_p017_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Garden path effects and correlations from predictions derived from Syntactic Belief Update in relation to [PITH_FULL_IMAGE:figures/full_fig_p018_6.png] view at source ↗
read the original abstract

Garden path sentences present a processing difficulty for humans -- the sentence prefix leads the listener towards one interpretation, until the listener hears a critical word that shows that the initial interpretation was wrong. Lexical surprisal, a measure that usually predicts sentence processing difficulty quite well, fails to provide good predictions for garden path sentences. We propose an alternative that actively predicts a probability distribution over syntactic trees (its syntactic belief) and updates that distribution after each new word. If a processor is led down a garden path, syntactic beliefs will be wrong and will require a large update at the critical word. The magnitude of the update is measured with a generalized R\'enyi divergence. Crucially, this metric is dependent on lexical items, but is fully independent of the probability of lexical items. This Syntactic Belief Update provides a better fit to the human reading time data on garden path sentences. This suggests a new research direction examining purely non-lexical alternatives to surprisal for psycholinguistics.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper claims that lexical surprisal fails to predict processing difficulty in garden-path sentences, and proposes instead a 'Syntactic Belief Update' metric: a processor maintains a probability distribution over syntactic trees and updates it word-by-word; the magnitude of the update at the disambiguating word, quantified by generalized Rényi divergence, indexes difficulty. The metric is asserted to depend on lexical items yet remain fully independent of lexical probabilities, and is claimed to provide a better fit to human reading-time data than surprisal, opening a research direction on non-lexical alternatives.

Significance. If the independence claim and the improved fit can be substantiated, the work would supply a concrete, falsifiable syntactic alternative to surprisal that isolates belief revision over tree distributions from lexical probability effects, potentially explaining why surprisal under-predicts garden-path costs and motivating new experimental tests that manipulate syntactic uncertainty while holding lexical probabilities constant.

major comments (2)
  1. [Abstract] Abstract: the central claim that the Rényi-divergence metric 'is fully independent of the probability of lexical items' while still depending on lexical items is load-bearing for the non-lexical-alternative thesis, yet no formal update rule, parser, or belief-maintenance procedure is supplied that achieves this separation; without it the distinction from surprisal cannot be verified.
  2. [Abstract] Abstract: the assertion that 'This Syntactic Belief Update provides a better fit to the human reading time data' is the primary empirical claim, but the abstract (and the visible manuscript) supplies neither the syntactic parser, the precise form of the tree distribution, the dataset, the statistical model, nor any baseline comparison, rendering the fit claim unverifiable.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their comments, which highlight the need for greater formal and empirical detail. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the Rényi-divergence metric 'is fully independent of the probability of lexical items' while still depending on lexical items is load-bearing for the non-lexical-alternative thesis, yet no formal update rule, parser, or belief-maintenance procedure is supplied that achieves this separation; without it the distinction from surprisal cannot be verified.

    Authors: We agree that the independence claim requires an explicit formalization to be verifiable. The revised manuscript will include a dedicated methods section specifying the syntactic parser, the precise belief-maintenance procedure over tree distributions, and the update rule that conditions on lexical items while computing generalized Rényi divergence solely over syntactic structure probabilities. This will demonstrate the separation from lexical surprisal. revision: yes

  2. Referee: [Abstract] Abstract: the assertion that 'This Syntactic Belief Update provides a better fit to the human reading time data' is the primary empirical claim, but the abstract (and the visible manuscript) supplies neither the syntactic parser, the precise form of the tree distribution, the dataset, the statistical model, nor any baseline comparison, rendering the fit claim unverifiable.

    Authors: We acknowledge that the current manuscript version does not supply the methodological details needed to evaluate the fit claim. The revision will add the parser specification, the exact form of the tree distribution, the reading-time dataset, the statistical model (including how comparisons are performed), and explicit baseline results against surprisal, allowing direct verification of the reported improvement. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation remains self-contained

full rationale

The paper asserts that a Rényi divergence over a syntactic belief distribution (updated word-by-word) is independent of lexical probabilities yet still depends on lexical items, and reports a better fit to garden-path reading times. No equations, parameter-fitting steps, or self-citations are exhibited that reduce the divergence metric by construction to quantities fitted on the target reading-time data. The independence claim is presented as a modeling choice rather than derived from prior self-citations or ansatzes that smuggle in the result. The reported better fit is a standard model-comparison outcome and does not constitute a 'prediction' that is statistically forced by the same inputs. The derivation chain is therefore not equivalent to its inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the generalized Rényi divergence and syntactic belief distribution are introduced without further decomposition.

pith-pipeline@v0.9.1-grok · 5704 in / 1020 out tokens · 30291 ms · 2026-06-26T04:35:09.426884+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

299 extracted references · 178 canonical work pages · 2 internal anchors

  1. [1]

    Cognitive Science , author =

    Modeling. Cognitive Science , author =. 2023 , keywords =. doi:10.1111/cogs.13312 , abstract =

    work page doi:10.1111/cogs.13312 2023

  2. [2]

    and Carpenter, Patricia A

    Just, Marcel A. and Carpenter, Patricia A. and Woolley, Jacqueline D. , year =. Paradigms and processes in reading comprehension , volume =. Journal of Experimental Psychology: General , publisher =. doi:10.1037/0096-3445.111.2.228 , abstract =

    work page doi:10.1037/0096-3445.111.2.228

  3. [3]

    The syntactic process , isbn =

    Steedman, Mark , year =. The syntactic process , isbn =. doi:https://doi.org/10.7551/mitpress/6591.001.0001 , language =

    work page doi:10.7551/mitpress/6591.001.0001

  4. [4]

    Journal of Psycholinguistic Research , author =

    Subject-. Journal of Psycholinguistic Research , author =. 1999 , keywords =. doi:10.1023/A:1023206208142 , abstract =

    work page doi:10.1023/a:1023206208142 1999

  5. [5]

    Society for Computation in Linguistics Volume 7 Issue 1 2024 , publisher =

    Ozaki, Satoru and De Santo, Aniello and Linzen, Tal and Dillon, Brian , month = jun, year =. Society for Computation in Linguistics Volume 7 Issue 1 2024 , publisher =. doi:10.7275/SCIL.2229 , abstract =

    work page doi:10.7275/scil.2229 2024

  6. [6]

    , year =

    Pritchett, Bradley L. , year =. Garden. Language , publisher =. doi:10.2307/414532 , abstract =

    work page doi:10.2307/414532

  7. [7]

    Cognitive Psychology , author =

    Making and correcting errors during sentence comprehension:. Cognitive Psychology , author =. 1982 , keywords =. doi:10.1016/0010-0285(82)90008-1 , language =

    work page doi:10.1016/0010-0285(82)90008-1 1982

  8. [8]

    Misanalysis and

    Keshev, Maayan , keywords =. Misanalysis and

  9. [9]

    Journal of Memory and Language , author =

    Recovery from misanalyses of garden-path sentences , volume =. Journal of Memory and Language , author =. 1991 , keywords =. doi:10.1016/0749-596X(91)90034-H , abstract =

    work page doi:10.1016/0749-596x(91)90034-h 1991

  10. [10]

    Nature Neuroscience , author =

    Shared computational principles for language processing in humans and deep language models , volume =. Nature Neuroscience , author =. 2022 , pages =. doi:10.1038/s41593-022-01026-4 , abstract =

    work page doi:10.1038/s41593-022-01026-4 2022

  11. [11]

    Proceedings of the Annual Meeting of the Cognitive Science Society , author =

    How much harder are hard garden-path sentences than easy ones? , volume =. Proceedings of the Annual Meeting of the Cognitive Science Society , author =

  12. [12]

    and Kanwisher, Nancy and Tenenbaum, Joshua B

    Schrimpf, Martin and Blank, Idan Asher and Tuckute, Greta and Kauf, Carina and Hosseini, Eghbal A. and Kanwisher, Nancy and Tenenbaum, Joshua B. and Fedorenko, Evelina , month = nov, year =. The neural architecture of language:. Proceedings of the National Academy of Sciences , publisher =. doi:10.1073/pnas.2105646118 , abstract =

    work page doi:10.1073/pnas.2105646118

  13. [13]

    and Jaeger, T

    Kuperberg, Gina R. and Jaeger, T. Florian , month = jan, year =. What do we mean by prediction in language comprehension? , volume =. Language, Cognition and Neuroscience , publisher =. doi:10.1080/23273798.2015.1102299 , abstract =

    work page doi:10.1080/23273798.2015.1102299 2015

  14. [14]

    Zeldes, Amir and Schneider, Nathan , editor =. Are. Proceedings of the. 2023 , keywords =

    2023

  15. [15]

    , month = feb, year =

    Dozat, Timothy and Manning, Christopher D. , month = feb, year =. Deep. International

  16. [16]

    Proceedings of the 16th

    Ahrenberg, Lars , editor =. Proceedings of the 16th. 2007 , pages =

    2007

  17. [17]

    On comprehending sentences:

    Frazier, Lyn , year =. On comprehending sentences:

  18. [18]

    2009 , month =

    Bever, Thomas G. , editor =. The cognitive basis for linguistic structures , isbn =. Language. 2013 , keywords =. doi:10.1093/acprof:oso/9780199677139.003.0001 , urldate =

    work page doi:10.1093/acprof:oso/9780199677139.003.0001 2013

  19. [19]

    Wilcox, Ethan G and Gauthier, Jon and Hu, Jennifer and Qian, Peng and Levy, Roger P , keywords =. On the

  20. [20]

    Proceedings of the Annual Meeting of the Cognitive Science Society , author =

    Using the. Proceedings of the Annual Meeting of the Cognitive Science Society , author =. 2021 , keywords =

    2021

  21. [21]

    Wilcox, Ethan and Vani, Pranali and Levy, Roger , year =. A. Proceedings of the 59th. doi:10.18653/v1/2021.acl-long.76 , abstract =

    work page doi:10.18653/v1/2021.acl-long.76 2021

  22. [22]

    Journal of Experimental Psychology

    Rapid syntactic adaptation in self-paced reading:. Journal of Experimental Psychology. Learning, Memory, and Cognition , author =. 2021 , keywords =. doi:10.1037/xlm0001046 , abstract =

    work page doi:10.1037/xlm0001046 2021

  23. [23]

    Transactions of the Association for Computational Linguistics , author =

    Testing the. Transactions of the Association for Computational Linguistics , author =. 2023 , keywords =. doi:10.1162/tacl_a_00612 , abstract =

    work page doi:10.1162/tacl_a_00612 2023

  24. [24]

    A very basic tutorial for performing linear mixed effects for analyses , url =

    Winter, Bodo , keywords =. A very basic tutorial for performing linear mixed effects for analyses , url =

  25. [25]

    Language, Cognition and Neuroscience , author =

    On the parsing of garden-path sentences , volume =. Language, Cognition and Neuroscience , author =. 2021 , keywords =. doi:10.1080/23273798.2021.1922727 , abstract =

    work page doi:10.1080/23273798.2021.1922727 2021

  26. [26]

    1998 , keywords =

    Reanalysis in. 1998 , keywords =. doi:10.1007/978-94-015-9070-9 , language =

    work page doi:10.1007/978-94-015-9070-9 1998

  27. [27]

    Language and Linguistics Compass , author =

    Do. Language and Linguistics Compass , author =. 2008 , note =. doi:10.1111/j.1749-818X.2008.00071.x , abstract =

    work page doi:10.1111/j.1749-818x.2008.00071.x 2008

  28. [28]

    Computational

    Brasoveanu, Adrian and Dotlačil, Jakub , year =. Computational. doi:10.1007/978-3-030-31846-8 , language =

    work page doi:10.1007/978-3-030-31846-8

  29. [29]

    Timkey, William and Linzen, Tal , editor =. A. Findings of the. 2023 , keywords =. doi:10.18653/v1/2023.findings-emnlp.582 , abstract =

    work page doi:10.18653/v1/2023.findings-emnlp.582 2023

  30. [30]

    Comprehension effort as the cost of inference , abstract =

    Vigly, Jacob Hoover and Qian, Peng and Sonderegger, Morgan and O’Donnell, Timothy J , keywords =. Comprehension effort as the cost of inference , abstract =

  31. [31]

    Incremental

    Hanna, Michael and Mueller, Aaron , editor =. Incremental. Proceedings of the 2025. 2025 , keywords =. doi:10.18653/v1/2025.naacl-long.164 , abstract =

    work page doi:10.18653/v1/2025.naacl-long.164 2025

  32. [32]

    A probabilistic

    Hale, John , year =. A probabilistic. Second

  33. [33]

    , editor =

    Lewis, Richard L. , editor =. Reanalysis and. Reanalysis in. 1998 , note =. doi:10.1007/978-94-015-9070-9_8 , abstract =

    work page doi:10.1007/978-94-015-9070-9_8 1998

  34. [34]

    and Kemp, Charles and Tenenbaum, Joshua B

    Griffiths, Thomas L. and Kemp, Charles and Tenenbaum, Joshua B. , editor =. Bayesian. The. 2008 , keywords =. doi:10.1017/CBO9780511816772.006 , abstract =

    work page doi:10.1017/cbo9780511816772.006 2008

  35. [35]

    Proceedings of the Annual Meeting of the Cognitive Science Society , author =

    Modeling garden path effects without explicit hierarchical syntax , volume =. Proceedings of the Annual Meeting of the Cognitive Science Society , author =. 2018 , keywords =

    2018

  36. [36]

    beverpdf\_11 , url =

  37. [37]

    , year =

    Piantadosi, Steven T. , year =. Modern language models refute. From fieldwork to linguistic theory: A tribute to Dan Everett , publisher =

  38. [38]

    Narayanan, Srini and Jurafsky, Dan , editor =. A. Advances in. 2002 , keywords =. doi:10.7551/mitpress/1120.003.0012 , abstract =

    work page doi:10.7551/mitpress/1120.003.0012 2002

  39. [39]

    Bayesian

    Narayanan, Srini and Jurafsky, Daniel , collaborator =. Bayesian. Proceedings of the. 1998 , keywords =. doi:10.4324/9781315782416-138 , abstract =

    work page doi:10.4324/9781315782416-138 1998

  40. [40]

    Adopting

    Volenec, Veno and Reiss, Charles , keywords =. Adopting

  41. [41]

    Kajikawa, Kohei and Isono, Shinnosuke , keywords =. The

  42. [42]

    Clair, Cody and Fodor, Paul and Shibata, Chihiro and Heinz, Jeffrey , month = nov, year =

    van der Poel, Sam and Lambert, Dakotah and Kostyszyn, Kalina and Gao, Tiantian and Verma, Rahul and Andersen, Derek and Chau, Joanne and Peterson, Emily and St. Clair, Cody and Fodor, Paul and Shibata, Chihiro and Heinz, Jeffrey , month = nov, year =. doi:10.5061/DRYAD.DNCJSXM4H , abstract =

    work page doi:10.5061/dryad.dncjsxm4h

  43. [43]

    Bigelow, Eric and Wurgaft, Daniel and Wang, YingQiao and Goodman, Noah and Ullman, Tomer and Tanaka, Hidenori and Lubana, Ekdeep Singh , month = nov, year =. Belief. doi:10.48550/arXiv.2511.00617 , abstract =

    work page doi:10.48550/arxiv.2511.00617

  44. [44]

    Graph-based

    Swayamdipta, Swabha and Thomson, Sam , keywords =. Graph-based

  45. [45]

    Taskar, Ben and Klein, Dan and Collins, Michael and Koller, Daphne and Manning, Christopher , editor =. Max-. Proceedings of the 2004. 2004 , pages =

    2004

  46. [46]

    Thomas and Griffiths, Thomas L

    McCoy, R. Thomas and Griffiths, Thomas L. , month = may, year =. Modeling rapid language learning by distilling. Nature Communications , publisher =. doi:10.1038/s41467-025-59957-y , abstract =

    work page doi:10.1038/s41467-025-59957-y

  47. [47]

    Constraints on

    COWPER, ELIZABETH ANN , year =. Constraints on

  48. [48]

    An architecturally-based theory of human sentence comprehension , copyright =

    Lewis, Richard Lawrence , year =. An architecturally-based theory of human sentence comprehension , copyright =

  49. [49]

    Gibson, Edward , keywords =. Syntax:

  50. [50]

    Holographic

    Yamaki, Ryosuke and Taniguchi, Tadahiro and Mochihashi, Daichi , editor =. Holographic. Proceedings of the 61st. 2023 , keywords =. doi:10.18653/v1/2023.acl-long.15 , abstract =

    work page doi:10.18653/v1/2023.acl-long.15 2023

  51. [51]

    Holographic

    Higashi, Noji and Japan, Shiga , keywords =. Holographic

  52. [52]

    Ma, Xuezhe and Hovy, Eduard , editor =. Neural. Proceedings of the. 2017 , keywords =

    2017

  53. [53]

    Eelbrain, a

    Brodbeck, Christian and Das, Proloy and Gillis, Marlies and Kulasingham, Joshua P and Bhattasali, Shohini and Gaston, Phoebe and Resnik, Philip and Simon, Jonathan Z , editor =. Eelbrain, a. eLife , publisher =. 2023 , keywords =. doi:10.7554/eLife.85012 , abstract =

    work page doi:10.7554/elife.85012 2023

  54. [54]

    A tradeoff between acoustic and linguistic feature encoding in spoken language comprehension , volume =

    Tezcan, Filiz and Weissbart, Hugo and Martin, Andrea E , editor =. A tradeoff between acoustic and linguistic feature encoding in spoken language comprehension , volume =. eLife , publisher =. 2023 , keywords =. doi:10.7554/eLife.82386 , abstract =

    work page doi:10.7554/elife.82386 2023

  55. [55]

    Journal of Statistical Computation and Simulation , author =

    Elements of causal inference: foundations and learning algorithms , volume =. Journal of Statistical Computation and Simulation , author =. 2018 , keywords =. doi:10.1080/00949655.2018.1505197 , language =

    work page doi:10.1080/00949655.2018.1505197 2018

  56. [56]

    Statistics Surveys , author =

    Causal inference in statistics:. Statistics Surveys , author =. 2009 , keywords =. doi:10.1214/09-SS057 , abstract =

    work page doi:10.1214/09-ss057 2009

  57. [57]

    Unzipping the

    Doucette, Amanda and O’Donnell, Timothy J and Sonderegger, Morgan , keywords =. Unzipping the

  58. [58]

    Linguistic Variation Yearbook , author =

    Towards a restrictive theory of (remnant) movement! , volume =. Linguistic Variation Yearbook , author =. 2007 , keywords =. doi:10.1075/livy.7.04abe , abstract =

    work page doi:10.1075/livy.7.04abe 2007

  59. [59]

    Incremental

    Hanna, Michael and Mueller, Aaron , month = dec, year =. Incremental. doi:10.48550/arXiv.2412.05353 , abstract =

    work page doi:10.48550/arxiv.2412.05353

  60. [60]

    Computational Linguistics , author =

    Are. Computational Linguistics , author =. 2025 , pages =. doi:10.1162/coli.a.24 , abstract =

    work page doi:10.1162/coli.a.24 2025

  61. [61]

    Nivre, Joakim , editor =. Non-. Proceedings of the. 2009 , pages =

    2009

  62. [62]

    Language and Linguistics Compass , author =

    Antisymmetry and. Language and Linguistics Compass , author =. 2012 , pages =. doi:10.1002/lnc3.319 , abstract =

    work page doi:10.1002/lnc3.319 2012

  63. [63]

    Anticipation of clause-final heads

    Konieczny, Lars and Döring, Philipp , keywords =. Anticipation of clause-final heads

  64. [64]

    Journal of Memory and Language , author =

    The syntactic complexity of. Journal of Memory and Language , author =. 2013 , keywords =. doi:10.1016/j.jml.2012.10.005 , abstract =

    work page doi:10.1016/j.jml.2012.10.005 2013

  65. [65]

    Nathan and Brennan, Jonathan R

    Li, Jixing and Bhattasali, Shohini and Zhang, Shulin and Franzluebbers, Berta and Luh, Wen-Ming and Spreng, R. Nathan and Brennan, Jonathan R. and Yang, Yiming and Pallier, Christophe and Hale, John , month = aug, year =. Le. Scientific Data , publisher =. doi:10.1038/s41597-022-01625-7 , abstract =

    work page doi:10.1038/s41597-022-01625-7

  66. [66]

    , year =

    Hale, John T. , year =. Automaton theories of human sentence comprehension , isbn =

  67. [67]

    Behavior Research Methods , author =

    Expanding horizons of cross-linguistic research on reading:. Behavior Research Methods , author =. 2022 , keywords =. doi:10.3758/s13428-021-01772-6 , abstract =

    work page doi:10.3758/s13428-021-01772-6 2022

  68. [68]

    and Brasser, Jan and Campos-Rojas, César and Drieghe, Denis and Filipović Đurđević, Dušica and Goldina, Sofya and Ibáñez Orellana, Romualdo and Jäger, Lena A

    Siegelman, Noam and Schroeder, Sascha and Bao, Yaqian Borogjoon and Acartürk, Cengiz and Agrawal, Niket and Bolliger, Lena S. and Brasser, Jan and Campos-Rojas, César and Drieghe, Denis and Filipović Đurđević, Dušica and Goldina, Sofya and Ibáñez Orellana, Romualdo and Jäger, Lena A. and Jóhannesson, Ómar I. and Khare, Anurag and Kharlamov, Nik and Knudse...

    work page doi:10.1038/s41597-025-05453-3

  69. [69]

    Journal of Psycholinguistic Research , author =

    Against. Journal of Psycholinguistic Research , author =. 2003 , keywords =. doi:10.1023/A:1022496223965 , abstract =

    work page doi:10.1023/a:1022496223965 2003

  70. [70]

    Journal of Memory and Language , author =

    Structural. Journal of Memory and Language , author =. 1999 , pages =. doi:10.1006/jmla.1998.2606 , language =

    work page doi:10.1006/jmla.1998.2606 1999

  71. [71]

    Cognitive Science , author =

    Consequences of the. Cognitive Science , author =. 2005 , pages =. doi:10.1207/s15516709cog0000_7 , abstract =

    work page doi:10.1207/s15516709cog0000_7 2005

  72. [72]

    Sha, Fei and Pereira, Fernando , year =. Shallow. Proceedings of the 2003

    2003

  73. [73]

    Proceedings of the 2020

    Pfeiffer, Jonas and Rücklé, Andreas and Poth, Clifton and Kamath, Aishwarya and Vulić, Ivan and Ruder, Sebastian and Cho, Kyunghyun and Gurevych, Iryna , editor =. Proceedings of the 2020. 2020 , keywords =. doi:10.18653/v1/2020.emnlp-demos.7 , abstract =

    work page doi:10.18653/v1/2020.emnlp-demos.7 2020

  74. [74]

    Goldfish: Monolingual Language Models for 350 Languages

    Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K. , month = aug, year =. Goldfish:. doi:10.48550/arXiv.2408.10441 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2408.10441

  75. [75]

    , year =

    Kayne, Richard S. , year =. The antisymmetry of syntax , isbn =

  76. [76]

    Theoretical Linguistics , author =

    Cyclic. Theoretical Linguistics , author =. 2005 , keywords =. doi:10.1515/thli.2005.31.1-2.1 , abstract =

    work page doi:10.1515/thli.2005.31.1-2.1 2005

  77. [77]

    HuggingFace's Transformers: State-of-the-art Natural Language Processing

    Wolf, Thomas and Debut, Lysandre and Sanh, Victor and Chaumond, Julien and Delangue, Clement and Moi, Anthony and Cistac, Pierric and Rault, Tim and Louf, Rémi and Funtowicz, Morgan and Davison, Joe and Shleifer, Sam and Platen, Patrick von and Ma, Clara and Jernite, Yacine and Plu, Julien and Xu, Canwen and Scao, Teven Le and Gugger, Sylvain and Drame, M...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1910.03771 1910

  78. [78]

    Universal

    Nivre, Joakim and Zeman, Daniel and Ginter, Filip and Tyers, Francis , editor =. Universal. Proceedings of the 15th. 2017 , keywords =

    2017

  79. [79]

    Hierarchical syntactic structure in human-like language models , url =

    Wolfman, Michael and Dunagan, Donald and Brennan, Jonathan and Hale, John , editor =. Hierarchical syntactic structure in human-like language models , url =. Proceedings of the. 2024 , keywords =. doi:10.18653/v1/2024.cmcl-1.6 , abstract =

    work page doi:10.18653/v1/2024.cmcl-1.6 2024

  80. [80]

    Brain and Language , author =

    Abstract linguistic structure correlates with temporal activity during naturalistic comprehension , volume =. Brain and Language , author =. 2016 , keywords =. doi:10.1016/j.bandl.2016.04.008 , abstract =

    work page doi:10.1016/j.bandl.2016.04.008 2016

Showing first 80 references.