{"id":"W2116795926","doi":"10.1098/rspb.2004.2957","title":"The consequences of Zipf's law for syntax and symbolic reference","year":2005,"lang":"en","type":"article","venue":"Proceedings of the Royal Society B Biological Sciences","topic":"Language and cultural evolution","field":"Social Sciences","cited_by":114,"is_retracted":false,"has_abstract":true,"ca_institutions":"Trinity College","funders":"Defense Advanced Research Projects Agency; Generalitat de Catalunya; Royal Society; National Science Foundation","keywords":"Zipf's law; Syntax; Ambiguity; Computer science; sort; Communications system; Entropy (arrow of time); Theoretical computer science; Natural language processing; Programming language; Mathematics; Telecommunications","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":["sts"],"category_scores_codex":[0.001445066,0.00007364127,0.0001265902,0.000003396605,0.001721595,0.00007859264,0.0006199435,0.0001015752,0.00001600076],"category_scores_gemma":[0.0004554054,0.00002934362,0.0001181499,0.0002133332,0.006625501,0.0001348787,0.0001069561,0.0000745881,6.889902e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002624887,"about_ca_system_score_gemma":0.00004456662,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001105,"about_ca_topic_score_gemma":0.0001250323,"domain_scores_codex":[0.9990804,0.00002143055,0.0001790657,0.0001833857,0.000272162,0.0002635316],"domain_scores_gemma":[0.9992874,0.0002626918,0.0001981507,0.00003150771,0.0001749518,0.00004531563],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"qualitative","study_design_scores_codex":[0.00001670464,0.00003013645,0.01555468,0.00001928442,0.0000174189,7.5659e-9,0.005976751,0.000002595496,0.02063763,0.9507107,0.001503424,0.005530661],"study_design_scores_gemma":[0.00116584,0.001904319,0.07065477,0.0004236373,0.000182371,0.000006208623,0.3022344,0.001552453,0.1411027,0.3014933,0.1779987,0.001281279],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9727543,0.001058609,0.000002650804,0.006319278,0.0000515531,0.0002758537,0.000007007489,0.00002144175,0.0195093],"genre_scores_gemma":[0.9981355,0.0003917098,0.0007426409,0.0002398131,0.0001041298,0.00001786734,1.761971e-7,0.000001081096,0.0003670624],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6492174,"threshold_uncertainty_score":0.999578,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04295064284193948,"score_gpt":0.3025182159496375,"score_spread":0.259567573107698,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}