FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2767833864> ?p ?o ?g. }

• W2767833864 abstract "From Syllables to Syntax: Investigating Staged Linguistic Development through Computational Modeling Kris Jack (kjack@computing.dundee.ac.uk) Chris Reed (chris@computing.dundee.ac.uk) Annalu Waller (awaller@computing.dundee.ac.uk) Applied Computing, University of Dundee, Dundee, DD1 4HN, Scotland. Abstract Modeling the Emergence of Syntax All natural languages employ syntax. Syntax allows individuals to both understand and produce novel utterances. Unlike non-formulaic language, syntactically produced utterances are a function of their internal parts. Elman (1993) finds that simple and complex linguistic structures can be learned by a neural network, but only if the former are acquired before the latter. To ensure simple structures are learned first, the neural network's memory length is initially small, and increased during training. This 'maturational' growth allows both types to be acquired without staged input. Dominey and Boucher (2005) investigate developmental phenomena within a grounded robot. A form of syntactic bootstrapping arises as grounded pairs are learned. The model, however, employs a manual trigger that activates the syntactic component, an inadequate explanation for the emergence of syntax. Kirby (2001) considers language transmission from generation to generation through the Iterated Learning Model. He demonstrates that transmission bottlenecks, that determine the amount of linguistic exposure a learner receives, have an important effect on the emergence of syntax. The bottleneck can be neither too narrow nor too wide for syntactic structures to be derived. A new model of early language acquisition is introduced. The model demonstrates the staged emergence of lexical and syntactic acquisition. For a period, no linguistic activity is present. The emergence of first words signals the onset of the holophrastic stage that continues to mature without syntactic activity. Syntactic awareness eventually emerges as the result of multiple lexically-based insights. No mechanistic triggers are employed throughout development. Keywords: Computational modeling; Emergence of Syntax; Item-based Learning; Language Acquisition. Introduction Children acquire language in stages, first learning words and later showing sensitivity to their syntactic properties. Processes that demonstrate distinct behaviors at different stages of development are difficult to model within a unified system. As a result, lexical and syntactic processes are often modeled independently from one another. Bridging the gap between these models will increase understanding of the behavioral shift that ushers in syntactic awareness. Background Modeling Word-to-meaning Mappings Bridging the Gap between Words and Syntax Children learn the meanings of a small number of words early in linguistic development. These first words are often non-formulaic (Wray, 2002). A non-formulaic word expresses a word-to-meaning relationship that is not a function of the word's internal parts. Siskind (1996) investigates word-to-meaning mappings using cross-situational analysis. Cross-situational analysis takes advantage of word-meaning co-occurrences to establish relationships. His simulations show considerable success, offering a robust solution to the problem under a variety of circumstances. Steels (2001) considers the problem of establishing such mappings through language games. Treating language as a complex adaptive system, he shows that social pressures to communicate, through games, encourage the development of a self-organized lexicon. Lexical acquisition is also studied within a developmental framework. Regier (2005) shows that interesting lexical phenomena, such as fast-mapping, can arise without internal mechanistic changes. Attentional learning plays an important role in language acquisition. None of these models show the developmental shift from lexical to syntactic awareness reflected in child language development. Jack, Reed and Waller (2004) consider the transition from the one-word stage to the two-word stage. A model is trained on pairs, testing interpretation of strings at each training epoch. In early training, a preference for non-formulaic (lexical) interpretation emerges. As training continues, this preference fades, giving way to formulaic (syntactic) interpretations. The behavioral change is an emergent property of the training process and not artificially triggered. Although a developmental shift is witnessed it appears very early in the model and the purely lexical period is very short, unreflective of natural child language development. Modeling the Developmental Shift Children do not understand syntactically complex utterances from birth. First words, produced at around 10-months-old (Bates & Goodman, 1999), are non-formulaic, with no indication of syntactic properties. By around 18-months- old, syntactic awareness emerges (MacWhinney & Bates," @default.
• W2767833864 created "2017-11-17" @default.
• W2767833864 creator A5009492024 @default.
• W2767833864 creator A5016589971 @default.
• W2767833864 creator A5073692224 @default.
• W2767833864 date "2006-01-01" @default.
• W2767833864 modified "2023-09-28" @default.
• W2767833864 title "From Syllables to Syntax: Investigating Staged Linguistic Development through Computational Modeling" @default.
• W2767833864 cites W1502101714 @default.
• W2767833864 cites W1597524164 @default.
• W2767833864 cites W189137083 @default.
• W2767833864 cites W1952254 @default.
• W2767833864 cites W1999834439 @default.
• W2767833864 cites W2030527368 @default.
• W2767833864 cites W2033943155 @default.
• W2767833864 cites W2071402670 @default.
• W2767833864 cites W2093221030 @default.
• W2767833864 cites W2103064571 @default.
• W2767833864 cites W2147587545 @default.
• W2767833864 cites W2150961342 @default.
• W2767833864 cites W2151834591 @default.
• W2767833864 cites W2160783091 @default.
• W2767833864 cites W2171060340 @default.
• W2767833864 cites W2540180329 @default.
• W2767833864 cites W3087978717 @default.
• W2767833864 cites W35832374 @default.
• W2767833864 hasPublicationYear "2006" @default.
• W2767833864 type Work @default.
• W2767833864 sameAs 2767833864 @default.
• W2767833864 citedByCount "1" @default.
• W2767833864 crossrefType "journal-article" @default.
• W2767833864 hasAuthorship W2767833864A5009492024 @default.
• W2767833864 hasAuthorship W2767833864A5016589971 @default.
• W2767833864 hasAuthorship W2767833864A5073692224 @default.
• W2767833864 hasConcept C138885662 @default.
• W2767833864 hasConcept C149782125 @default.
• W2767833864 hasConcept C154945302 @default.
• W2767833864 hasConcept C186644900 @default.
• W2767833864 hasConcept C199360897 @default.
• W2767833864 hasConcept C204321447 @default.
• W2767833864 hasConcept C207609745 @default.
• W2767833864 hasConcept C33923547 @default.
• W2767833864 hasConcept C41008148 @default.
• W2767833864 hasConcept C41895202 @default.
• W2767833864 hasConcept C60048249 @default.
• W2767833864 hasConceptScore W2767833864C138885662 @default.
• W2767833864 hasConceptScore W2767833864C149782125 @default.
• W2767833864 hasConceptScore W2767833864C154945302 @default.
• W2767833864 hasConceptScore W2767833864C186644900 @default.
• W2767833864 hasConceptScore W2767833864C199360897 @default.
• W2767833864 hasConceptScore W2767833864C204321447 @default.
• W2767833864 hasConceptScore W2767833864C207609745 @default.
• W2767833864 hasConceptScore W2767833864C33923547 @default.
• W2767833864 hasConceptScore W2767833864C41008148 @default.
• W2767833864 hasConceptScore W2767833864C41895202 @default.
• W2767833864 hasConceptScore W2767833864C60048249 @default.
• W2767833864 hasIssue "28" @default.
• W2767833864 hasLocation W27678338641 @default.
• W2767833864 hasOpenAccess W2767833864 @default.
• W2767833864 hasPrimaryLocation W27678338641 @default.
• W2767833864 hasRelatedWork W1568142600 @default.
• W2767833864 hasRelatedWork W1792353940 @default.
• W2767833864 hasRelatedWork W1940261393 @default.
• W2767833864 hasRelatedWork W1982016222 @default.
• W2767833864 hasRelatedWork W200282793 @default.
• W2767833864 hasRelatedWork W2066286707 @default.
• W2767833864 hasRelatedWork W2228944440 @default.
• W2767833864 hasRelatedWork W2286453087 @default.
• W2767833864 hasRelatedWork W2415378728 @default.
• W2767833864 hasRelatedWork W2579840627 @default.
• W2767833864 hasRelatedWork W2745921773 @default.
• W2767833864 hasRelatedWork W2753257400 @default.
• W2767833864 hasRelatedWork W2767331001 @default.
• W2767833864 hasRelatedWork W2809590726 @default.
• W2767833864 hasRelatedWork W2951806831 @default.
• W2767833864 hasRelatedWork W3036699898 @default.
• W2767833864 hasRelatedWork W319306244 @default.
• W2767833864 hasRelatedWork W771671331 @default.
• W2767833864 hasRelatedWork W2022894875 @default.
• W2767833864 hasRelatedWork W2462659769 @default.
• W2767833864 hasVolume "28" @default.
• W2767833864 isParatext "false" @default.
• W2767833864 isRetracted "false" @default.
• W2767833864 magId "2767833864" @default.
• W2767833864 workType "article" @default.