<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div style="direction: ltr; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Dear Colleagues, </div>
<div style="direction: ltr; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
I am writing to ask for your help with the following issue. I have been working on the relationship between LLMs and linguistic theory and now it is time to check how language acquisition happens in humans and machines. I am therefore looking for the very first,
at least, 100 words of children with L1 English (in the ideal case, data should come from different varieties of the language). Currently, for technical reasons, LA comparisons are possible only for English. </div>
<div style="direction: ltr; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div class="elementToProof" style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
So far, I have worked with the following British English data: <span style="color: rgb(17, 85, 204);">
<u><a href="https://childes.talkbank.org/access/Eng-UK/Sekali.html" id="OWA7f5a32f3-7c9b-61ac-92e2-8d673c1e5d91" class="x_x_x_OWAAutoLink" data-auth="NotApplicable" style="color: rgb(17, 85, 204); margin: 0px;">https://childes.talkbank.org/access/Eng-UK/Sekali.html</a></u></span>.
However, it turned out that CHILDES is problematic in many respects: too old data, too late starting of the recordings, recording of the parents' input but not of the child's answers, especially at the early stages, etc. (I thank Katharina Korecky-Kröll for
invaluable help with CHILDES.) I therefore decided to turn to the <i>lingtyp </i>
community. Please feel free to forward the query to linguists that are not on the list. </div>
<div style="direction: ltr; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Thanks in advance for your help. If interesting things come out, I'll post a note. </div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
[Based on the CHILDES data mentioned above and a comparison with the ChatGPT vocabulary, LA of English in both humans and machines should happen in the same way. A short note on linguistic theory, psycholinguistics and LLMs, and how I compare things, at:
<span style="color: rgb(17, 85, 204);"><u><a href="https://lingbuzz.net/lingbuzz/008123" id="OWA97d031cb-cee0-8399-db53-ab3b9e25a118" class="x_x_x_OWAAutoLink" data-auth="NotApplicable" style="color: rgb(17, 85, 204); margin: 0px;">https://lingbuzz.net/lingbuzz/008123</a></u></span>.
Any criticism is welcome!]</div>
<div style="direction: ltr; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Best,</div>
<div style="direction: ltr; text-align: justify; line-height: 1.2; margin: 0pt 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Stela</div>
<div style="direction: ltr; margin: 0px; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div id="x_x_Signature">
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 16px; color: rgb(0, 0, 0);">
<span style="background-color: rgb(255, 255, 255);"><b>***</b></span></div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
<span style="background-color: rgb(255, 255, 255);"><b>Dr. Stela MANOVA, </b></span><b><i>Gauss:AI </i></b></div>
<div style="font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt; color: rgb(0, 0, 0);">
LingTransformer<sup>1,2</sup> / LearningTransformer / CodeTransformer</div>
<div style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt;">
<span style="color: rgb(0, 0, 0);">Email (default): </span><span style="color: rgb(12, 100, 192);"><a href="mailto:manova.stela@gmail.com" id="OWA74a3e73a-60ab-addb-cc9a-83182b50be3d" class="x_x_OWAAutoLink" title="mailto:manova.stela@gmail.com" style="text-decoration: none;">manova.stela@gmail.com</a></span></div>
<div style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt;">
<span style="color: rgb(0, 0, 0);">Email (alternative): </span><span style="color: rgb(12, 100, 192);"><a href="mailto:stela.manova@proton.me" id="OWA9e4c5b19-2630-f302-4c5f-55689e394be9" class="x_x_OWAAutoLink" title="mailto:stela.manova@proton.me" style="text-decoration: none;">manova.stela@proton.me</a> </span></div>
<div style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 11pt;">
<span style="color: rgb(0, 0, 0);">Web: </span><span style="color: rgb(12, 100, 192);"><a href="https://sites.google.com/view/stelamanova" id="OWAf9299377-beca-c2f5-2033-e15baaf2ee50" class="x_x_OWAAutoLink" title="https://sites.google.com/view/stelamanova" data-auth="NotApplicable" style="text-decoration: none;">https://sites.google.com/view/stelamanova</a></span></div>
<div style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 10pt; color: rgb(0, 0, 0);">
---</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 10pt;">
<span style="color: rgb(26, 26, 26);"><sup>1 </sup>"A <b><a href="https://en.wikipedia.org/wiki/Transformer_(deep_learning_architecture)" id="OWA0626d181-ca1c-620f-db6c-df7c7f364d27" class="x_x_OWAAutoLink" title="https://en.wikipedia.org/wiki/Transformer_(deep_learning_architecture)" data-auth="NotApplicable" style="text-decoration: none;">transformer</a></b> model
is a neural network that learns context and thus meaning by tracking relationships in sequential data like the words in this sentence.</span><span style="color: rgb(0, 0, 0);">",
</span><span style="color: rgb(12, 100, 192);"><a href="https://blogs.nvidia.com/blog/what-is-a-transformer-model" id="OWA6070b499-a3c3-9362-f347-fecadd721eec" class="x_x_OWAAutoLink" title="https://blogs.nvidia.com/blog/what-is-a-transformer-model/" data-auth="NotApplicable" style="text-decoration: none;">https://blogs.nvidia.com/blog/what-is-a-transformer-model</a></span><span style="color: rgb(0, 0, 0);">,
Wikipedia link inserted by SM.</span></div>
<div style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 10pt; color: rgb(12, 100, 192);">
---</div>
<div class="elementToProof" style="text-align: left; text-indent: 0px; background-color: rgb(255, 255, 255); margin: 0cm; font-family: Aptos, Aptos_EmbeddedFont, Aptos_MSFontService, Calibri, Helvetica, sans-serif; font-size: 10pt;">
<span style="color: rgb(0, 0, 0);"><sup>2</sup><b><sup> </sup>The LingTransformer:</b></span><span style="color: rgb(12, 100, 192);"> </span><span style="color: rgb(0, 0, 0);">I have a strong background in mathematics and computer science. I</span><span style="color: rgb(19, 19, 20); background-color: rgb(255, 255, 255);">n
2021, I openly claimed that syntactic trees used for formal representation of language (Chomsky's approach) are not hierarchical structures and have an unnatural direction of growth: from leaves to the root,
<a href="https://lingbuzz.net/lingbuzz/006082" id="OWAc2e9b6cc-b619-cc80-3630-dd804da85a88" class="x_x_OWAAutoLink" title="https://lingbuzz.net/lingbuzz/006082" data-auth="NotApplicable" style="text-decoration: none;">
lingbuzz/006082</a>, see also <a href="https://ling.auf.net/lingbuzz/007598" id="OWAa025c652-f9e1-983b-8b5a-199884236367" class="x_x_OWAAutoLink" title="https://ling.auf.net/lingbuzz/007598" data-auth="NotApplicable" style="text-decoration: none;">
lingbuzz/007598</a>. Since then, <a href="https://linguistics.mit.edu/user/chomsky/" id="OWA25a2482e-bb85-874e-8a67-d7287620994b" class="x_x_OWAAutoLink" title="https://linguistics.mit.edu/user/chomsky/" data-auth="NotApplicable" style="text-decoration: none;">
Noam Chomsky,</a> <a href="https://www.its.caltech.edu/~matilde/" id="OWA6e527f98-2994-81e2-fa28-7486e09315d5" class="x_x_OWAAutoLink" title="https://www.its.caltech.edu/~matilde/" data-auth="NotApplicable" style="text-decoration: none;">
Matilde Marcolli</a>, and <a href="https://idss.mit.edu/staff/robert-c-berwick/" id="OWAc7e0b865-11ef-817c-7f42-085a5502cf27" class="x_x_OWAAutoLink" title="https://idss.mit.edu/staff/robert-c-berwick/" data-auth="NotApplicable" style="text-decoration: none;">
Robert Berwick</a> have been looking for new representations of syntactic structures; for their papers, visit:
<a href="https://ling.auf.net/lingbuzz/_search?q=marcolli" id="OWAb0a31b60-0a9f-c1ca-9795-8e28e93ec9b3" class="x_x_OWAAutoLink" data-auth="NotApplicable" style="text-decoration: none;">
https://ling.auf.net/lingbuzz/_search?q=marcolli</a>. Curious why they will not succeed in their endeavor and what this means for research in linguistics, join
<b><i>Gauss:AI </i></b>(for the moment, just drop me an email message)!</span></div>
</div>
</body>
</html>