Thrilled to share a review on THE LANGUAGE NETWORK AS A NATURAL KIND—a culmination of ~20 yrs of thinking about+studying language from linguistic, psycholinguistic, and cog neuro perspectives. @NatRevNeurosci
With the amazing @neuranna @tamaregev 🥳
🧵1/n rdcu.be/dEylV
TLDR: The🧠language network is distinct from i)perceptual+premotor lang-relev. mechanisms (eg,speech perception+Broca’s articulation areas)—selective for ling inputs/outputs but insensitive to meaning; and ii)systems of thought—important for language use but not lang-specific.2/n

Thrilled to share this tour de force co-led by SammyFloyd+@OlessiaJour! 8yrs in the making! "A tripartite structure of pragmatic language abilities: comprehension of social conventions,intonation processing,and causal reasoning". W/@ZachMineroff; co-supervised w/@LanguageMIT 1/n Link to the preprint: .
Language comprehension requires pragmatic reasoning, which encompasses diverse phenomena: from irony, to conversational implicatures, to contrastive and emotional prosody, to indirect requests, etc. 2/npsyarxiv.com/e2xta

FFA: THAT a functionally specialized module for processing faces exists is not debated. A great summary of the history of the debate and the current state of affairs can be found here: jneurosci.org/content/37/5/1…
How this module arises remains an active area of research.

https://twitter.com/JamieReilly_cog/status/1639643481210515456

vWFA: It is not debated that in most proficient readers a highly specialized region exists on the bottom of the left ventral temporal lobe that responds to printed words and letter strings (not differentially). First evidence of robust response selectivity came from

What happens when parts of the🧠that eventually become a highly functionally specialized language system (left frontal and temporal areas) get damaged at birth or shortly thereafter? Where does the language system live in such brains? Long answer short: it's complicated! 1/n One claim (Lenneberg,1967) is that early left-hemisphere (LH) injury results in the lang system emerging in the right hemisphere (RH). But until recently, this claim could not be properly evaluated because indiv.-level functional mapping of speech+lang areas was not possible. 2/n

As promised, a🧵on language and thought. The relationship between language and thought has long been pondered and debated. It may be one of the deepest and most exciting questions in cognitive science. 1/n Ray Jackendoff sets up this question beautifully in his 1996 book (he proceeds to argue for the relative independence between the two). 2/n

New position paper on expressive agrammatism in aphasia (w/@raryskin,@LanguageMIT). Given growing evidence that humans approximate a Bayesian ideal observer, we revive the “economy of effort” idea where agrammatic output is a rational adaptation to increased production costs. 🧵 Paper link:
Speech (and written output) of many patients with non-fluent aphasia consists of short, simple utterances, with omitted/incorrect function words and morphology (e.g., number agreement). A video of a sample patient: psyarxiv.com/46vdu

Thx to the hard work of @ben_lipkin @GretaTuckute and others + @NIDCD funding, we are ready to present to you LanA (Language Atlas)—a probabilistic atlas for the lang network based on lang localizer data from >800 inds (available for both volume and surface brain spaces). 1/n The 1st subj in this dataset was scanned on Sept 4 ‘07. I had defended my thesis just 3 mo earlier and was starting to fully dive into the brain & lang work with @Nancy_Kanwisher. The last subj was scanned in June ‘21. So, these data have taken us ~14 years to amass. 2/n

Psyched to share the first results from a ~7yr journey, from an email I sent to @Nancy_Kanwisher in 2014 about an idea to develop lang(uage) 'localizers' for as many of the world's langs as possible to having fMRI data from native speakers of 45 langs. tinyurl.com/b5xv7e32 1/n The effort was co-led by the awesome @saima_mm and Dima Ayyash, with help from @Jeanne1Gallee (who laid the critical foundation for this project back in 2015) @AffourtitJosef MalteHoffman @ZachMineroff and @OlessiaJour along with many others (see our long Acknowledgments!). 2/n

Excited to share a new preprint! Bottom line: the system that supports language processing (including syntactic processing) does not process music (including music structure): tinyurl.com/22yfyzcu This work was co-led by @Jessica_ChenXY and @AffourtitJosef Details below. 1/n Many *love* the idea that language and music draw on the same hierarchical-structure processor. I used to like it, too. BUT: i) no compelling evidence to date supports this idea; and ii) the idea is conceptually flawed. So, we decided to take another careful look. 2/n

This term, I taught my 1st class at MIT, "Language in the mind and brain". I have ideas for how to improve it next year and plan to make next year's lectures available online. For now, I wanted to share the art I used for the title slides because I had so much fun choosing these! Perceptual and motor foundations of language and introduction to high-level language processing:

In 2016, a woman (age 54 at the time) contacted us who was living without her left temporal lobe. She didn't know about her missing temp. lobe until age 25. She never experienced any head trauma/injury, so the temp. lobe was likely lost as a result of pre/perinatal stroke. 1/n As with most cases of early brain damage, she had no linguistic or cognitive deficits, but brains like EG's (fake initials) are invaluable for understanding how cognitive functions reorganize in the tissue that remains. I told her we definitely want to study her brain. 2/n

Inter-individual variability can cause two kinds of problems (in both lesion and brain-imaging analyses): missing effects that are there, and conflating nearby functional areas (the loss of 'functional resolution' as discussed in Nieto-Castañón & Fedorenko, 2012). 1/n

https://twitter.com/wmatchin/status/1311803481217523717

The latter is most relevant to your question: because lang areas in both frontal and temporal cortex lay adjacent to numerous functionally distinct areas, and precisely because lesions do not respect functional boundaries, any given lesion is likely to affect multiple areas. 2/n

Reduced language lateralization is one of the few replicated functional brain differences between individuals with autism spectrum disorders (ASD) and
neurotypical controls. However, past work has limitations and has left some key questions open:

https://twitter.com/OlessiaJour/status/1306240239170711557

#1: use of group analyses (=> Apparent reduction in activity in some brain areas in ASD at the group level may simply reflect higher variability in the locations of the functional regions)

This project began back in 2008, in my 1st year of postdoc'ing with @Nancy_Kanwisher. @spiantado, @LanguageMIT, and I worked on the stories (with Stacey Vishnevetsky, a UROP at that time, now a doctor-the real kind), and then @LanguageMIT and @Nancy_Kanwisher recorded these. :)

https://twitter.com/LanguageMIT/status/1303393597241405440

This paper releases the audio recordings, the SPR data, and some annotations. We also have eye-tracking data (with @tmalsburg and @roger_p_levy) and a bunch of fMRI data, all to be released soon.

Exciting to finally have this out! It has been surprisingly (to me) challenging to publish this paper in spite of clear results and the fact that this is the 1st functional neuroimaging investigation of this cool population. But better late than never. :) Congrats @OlessiaJour!

https://twitter.com/OlessiaJour/status/1296904145807515650

Population: 17 polyglots, all native speakers of English; median number of langs spoken with some level of proficiency =7 (mean =13.9, range: 5–55!). 9/17 polyglots qualified as “hyper-polyglots,” having some knowledge of 10 or more languages (Erard, 2012).

Exciting new work led by the amazing @martin_schrimpf (with @IbanDlank @GretaTuckute @KaufCarina EghbalHosseini @Nancy_Kanwisher JoshTenenbaum and me; @mitbrainandcog @MIT_CBMM @mcgovernmit): biorxiv.org/content/10.110… Inspired by the success of artificial neural networks (ANNs) in explaining neural responses in perceptual tasks, we tested 43 SOTA language models spanning major model classes (embedding models, recurrent networks, transformers) against three neural datasets (2 fMRI, 1 ECoG).

Thread: "Lack of selectivity for syntax relative to word meanings throughout the language network” tinyurl.com/y77cuznv is out in Cognition and is a culmination of one line of work, which I have been pursuing for the last ~15 years! Here is the story of how it all unfolded. with @IbanDlank @ZachMineroff @MattSiegelman

Now online at JNeurosci: The domain-general multiple demand (MD) network does not support core aspects of language comprehension: a large-scale fMRI investigation. Co-led by @ediachek, @IdanAsherBlank, and @MattSiegelman, also with @AffourtitJosef. [thread] Lexical access and synt/sem parsing are often described using domain-general cognitive constructs, like storing information in + retrieving it from WM, updating focal attention, inhibiting irrelevant information, selecting an option among alternatives, and predictive processing.

Excited to share a new preprint: biorxiv.org/content/10.110…
with LeilaWehbe @IdanAsherBlank @coryshain @rljfutrell @roger_p_levy @tmalsburg @LanguageMIT
"Incremental language comprehension difficulty predicts activity in the language network but not the multiple demand network" We compared fMRI data during naturalistic story listening to theory-neutral measures of online comprehension difficulty and incremental processing load: reading times in a self-paced reading paradigm and eye-fixation durations.

Now online: Broca's area is not a natural kind @TrendsCognSci with @IdanAsherBlank
doi.org/10.1016/j.tics…
Key message: Broca’s area, or its anatomical subdivisions (e.g., IFGop or IFGri), do not constitute meaningful units of analysis. 1/n Around 2008-10—as I was collecting and analyzing data for an fMRI study probing the selectivity of language-responsive areas —it became clear that right next to a language-selective region within Broca’s area lies a region with a very different functional profile. 2/n

Very exited about this work led by Cory Shain and @IdanAsherBlank, also with @marty_with_an_e and William Schuler:
biorxiv.org/content/10.110…
" fMRI reveals language-specific predictive coding during naturalistic sentence comprehension" 1/4 Using naturalistic materials and a large sample (n=78), we show that predictive coding in the brain’s response to language is implemented in domain-specific circuits, in line with the view of prediction as a "canonical computation" (Keller & Mrsic-Flogel, 2018). 2/4