@bergelsonlab
@bergelsonlab
Followers
2K
Following
5K
Media
324
Statuses
3K
Official account for the Bergelson Lab @HarvardU, sporadically maintained by the PI:)
Cambridge, MA
Joined April 2018
One of the only reason to hop back on X for a sec: promoting trainee work:)! Nice paper from @Fede_Bulgarelli taking a step towards mechanisms of early world learning!
New paper out in Developmental Science with @bergelsonlab! We asked whether the acoustic properties of language input for individual words also predict age of production. TLDR: Yes – variability in how words sound predicts when kids first say them. https://t.co/KxDcjV21uu (1/n)
0
0
1
If you like my work and think we could learn from each other and have fun doing research together, Marie Curie postdoctoral applications will soon open and my university has wonderful support for writing one. Write me a mail explaining your idea and why we'd match! 1/
1
20
48
Huge congratulations to @asifa_majid, recipient of the 2024 Jeffrey L. Elman Prize for Scientific Achievement and Community Building! This honor will be celebrated at #CogSci2024 with a prize and dedicated symposium. Visit the #ElmanPrize page for more https://t.co/N1tXSLx4az
14
28
139
:) big thanks to @christydesmith for being a wonderful journalist to work with, it was fun to chat with her for this profile!
Why do some kids learn to talk earlier than others? — Profile of my colleague Ellika Bergelson and her new findings and methods on language acquisition. https://t.co/esgdveTvSq)
0
0
5
“One really important shift in the field recently has been a much more serious reckoning with the fact that we tend to study white, middle-class Americans.” Important new insights into early childhood language learning, thanks to @bergelsonlab
https://t.co/YnUNt2HsA3
neurosciencenews.com
Findings reveal early comprehension begins around 6-7 months, and significant improvements in language understanding occur around a child’s first birthday.
1
3
7
Mieux comprendre l'écoute et la production du langage chez l'enfant. https://t.co/QSroHmeILH La quantité de paroles d'adultes entendues par un enfant aurait une influence significative sur le développement précoce du langage, l'effet augmentant avec l'âge. @CNRSshs @CNRS
0
5
7
Language development in children. The amount of adult speech a child hears has a significant additional factor in children speech production - @ENS_ULM @CristiaAlex @bergelsonlab @HarvardU @CNRS
Mieux comprendre l'écoute et la production du langage chez l'enfant. https://t.co/QSroHmeILH La quantité de paroles d'adultes entendues par un enfant aurait une influence significative sur le développement précoce du langage, l'effet augmentant avec l'âge. @CNRSshs @CNRS
0
1
4
nice coverage of our work! fun for me & @CristiaAlex to chat w/ @cathleenogrady &thnx to @oedemir & @mcxfrank for weighing in🤓. Read the whole paper to get our in depth take if u wanna know more: there’s a lot of nuance in these big topics that’s hard to relay in short #scicomm
A new analysis suggests kids speak more when the adults around them are more talkative, which may also give them a larger vocabulary early in life.
0
2
9
Roots! I "entered" social science as an anthropology major 35 years ago with an interest in how meanings were organized across cultures, and finally published a paper that provides new insight led by the amazing @mollyllewis in PNAS. Preprint: https://t.co/hU1LZH4tSk
2
12
50
Okay all thanks to @bergelsonlab I'm now on blusky too. https://t.co/2FxinAFovv. Happy to meet you there too.
0
1
3
Coda: tagging coauth’s whose handles i know on this crumbling platform: @CristiaAlex @CaroRowland @MarinaKalashnik + huge thanks to super generous data stewards & again to rad coauthors!🙏
1
0
2
Best for last: the #OpenScience goods: preprint: https://t.co/g6nEm74vq4, all code & de-id data for all #s in paper: https://t.co/BGdFqsuHjh; process to request data access: https://t.co/hxkzU4mooh 14/14
gin.g-node.org
Modern Research Data Management for Neuroscience
1
0
1
Point 4, applied:⚠️this isn’t an RCT & we make no causal claims⚠️. We found that for predicting everyday early👶 speech, amt of talk kids heard mattered, but mom’s ed didn’t. Speculation is cheap, causation is hard🤔. 13/14
1
0
1
Point 3, inclusivity: daylong recordings of the speech kids hear hold a lot of promise: They’re not as socioculturally biased as, say, standardized tests or lab sessions. They do wholly miss out on some stuff, (e.g. signed languages), leaving gaping holes worth pursuing 🕳️ 12/14
1
1
4
Point 2, methods: “big data” can help us understand human experience alongside detailed analysis of linguistic content (which we absolutely need, & which relies on hard-won manual transcription). As 🤖get better, we’ll gain more insights, e.g. about 🗣 content. 11/14
1
0
1
What’d we learn? Point 1, theory: obvi, 👶factors like age should predict👶🗣. Less obvi that enviro factors like amt. of adult 🗣 should. In fact, it runs counter to some big claims ab how we learn lang. Daily life is both super varied around the 🌍 & super consistent.10/14
1
0
3
(Isn't there more to 🗣 than QUANTITY?! Yes. But quant correlates strongly w/qualitative lang measures: hard to🗣 more w/o saying more stuff. More technically: amt of speech correlates w/word types, grammar complexity, etc. in natural 🗣. Less technically: bla bla bla ✅) 9/14
1
0
3
So, what predicted 👶🗣? Unsurprising: age & normative dev. More surprising: adult 🗣had about same effect as norm dev (for TOTALLY separate reasons). Other vars: ❎nada❎. Reviewers really wanted us to be sure mom’s ed didn’t matter. We looked. In 12 ways. It didn’t. 8/14
1
0
6
Prior work led us to consider some other variables too: amount of 🗣 kids heard from adults, mom’s education, child gender, & whether the kid was learning 1 or >1 lang. 7/14
1
0
2
We didn’t just wanna know what predicted 👶🗣, we wanted to know how much of an effect each variable had relative to others, in a more🌎 dataset Some variables we knew would matter (age & “normative” dev., i.e. diagnosis or high risk for a lang-related delay or not) 6/14
1
0
3