Hierarchical dynamic coding coordinates speech comprehension in the human brain

Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how rapid incoming sequences of hierarchical featur...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings of the National Academy of Sciences - PNAS Vol. 122; no. 42; p. e2422097122
Main Authors: Gwilliams, Laura, Marantz, Alec, Poeppel, David, King, Jean-Rémi
Format: Journal Article
Language:English
Published: United States 21.10.2025
Subjects:
ISSN:1091-6490, 1091-6490
Online Access:Get more information
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how rapid incoming sequences of hierarchical features are continuously coordinated. Here, we propose that each language feature is supported by a dynamic neural code, which represents the sequence history of hierarchical features in parallel. To test this "hierarchical dynamic coding" (HDC) hypothesis, we use time-resolved decoding of brain activity to track the construction, maintenance, and update of a comprehensive hierarchy of language features spanning phonetic, word form, lexical-syntactic, syntactic, and semantic representations. For this, we recorded 21 native English participants with magnetoencephalography (MEG), while they listened to two hours of short stories in English. Our analyses reveal three main findings. First, the brain represents and simultaneously maintains a sequence of hierarchical features. Second, the duration of these representations depends on their level in the language hierarchy. Third, each representation is maintained by a dynamic neural code, which evolves at a speed commensurate with its corresponding linguistic level. This HDC preserves the maintenance of information over time while limiting destructive interference between successive features. Overall, HDC reveals how the human brain maintains and updates the continuously unfolding language hierarchy during natural speech comprehension, thereby anchoring linguistic theories to their biological implementations.
AbstractList Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how rapid incoming sequences of hierarchical features are continuously coordinated. Here, we propose that each language feature is supported by a dynamic neural code, which represents the sequence history of hierarchical features in parallel. To test this "hierarchical dynamic coding" (HDC) hypothesis, we use time-resolved decoding of brain activity to track the construction, maintenance, and update of a comprehensive hierarchy of language features spanning phonetic, word form, lexical-syntactic, syntactic, and semantic representations. For this, we recorded 21 native English participants with magnetoencephalography (MEG), while they listened to two hours of short stories in English. Our analyses reveal three main findings. First, the brain represents and simultaneously maintains a sequence of hierarchical features. Second, the duration of these representations depends on their level in the language hierarchy. Third, each representation is maintained by a dynamic neural code, which evolves at a speed commensurate with its corresponding linguistic level. This HDC preserves the maintenance of information over time while limiting destructive interference between successive features. Overall, HDC reveals how the human brain maintains and updates the continuously unfolding language hierarchy during natural speech comprehension, thereby anchoring linguistic theories to their biological implementations.
Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how rapid incoming sequences of hierarchical features are continuously coordinated. Here, we propose that each language feature is supported by a dynamic neural code, which represents the sequence history of hierarchical features in parallel. To test this "hierarchical dynamic coding" (HDC) hypothesis, we use time-resolved decoding of brain activity to track the construction, maintenance, and update of a comprehensive hierarchy of language features spanning phonetic, word form, lexical-syntactic, syntactic, and semantic representations. For this, we recorded 21 native English participants with magnetoencephalography (MEG), while they listened to two hours of short stories in English. Our analyses reveal three main findings. First, the brain represents and simultaneously maintains a sequence of hierarchical features. Second, the duration of these representations depends on their level in the language hierarchy. Third, each representation is maintained by a dynamic neural code, which evolves at a speed commensurate with its corresponding linguistic level. This HDC preserves the maintenance of information over time while limiting destructive interference between successive features. Overall, HDC reveals how the human brain maintains and updates the continuously unfolding language hierarchy during natural speech comprehension, thereby anchoring linguistic theories to their biological implementations.Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the sensory input into increasingly abstract language properties. However, little is known about how rapid incoming sequences of hierarchical features are continuously coordinated. Here, we propose that each language feature is supported by a dynamic neural code, which represents the sequence history of hierarchical features in parallel. To test this "hierarchical dynamic coding" (HDC) hypothesis, we use time-resolved decoding of brain activity to track the construction, maintenance, and update of a comprehensive hierarchy of language features spanning phonetic, word form, lexical-syntactic, syntactic, and semantic representations. For this, we recorded 21 native English participants with magnetoencephalography (MEG), while they listened to two hours of short stories in English. Our analyses reveal three main findings. First, the brain represents and simultaneously maintains a sequence of hierarchical features. Second, the duration of these representations depends on their level in the language hierarchy. Third, each representation is maintained by a dynamic neural code, which evolves at a speed commensurate with its corresponding linguistic level. This HDC preserves the maintenance of information over time while limiting destructive interference between successive features. Overall, HDC reveals how the human brain maintains and updates the continuously unfolding language hierarchy during natural speech comprehension, thereby anchoring linguistic theories to their biological implementations.
Author King, Jean-Rémi
Gwilliams, Laura
Marantz, Alec
Poeppel, David
Author_xml – sequence: 1
  givenname: Laura
  orcidid: 0000-0002-9213-588X
  surname: Gwilliams
  fullname: Gwilliams, Laura
  organization: Stanford Data Science, Stanford University, Stanford, CA 94305
– sequence: 2
  givenname: Alec
  orcidid: 0000-0003-4805-2910
  surname: Marantz
  fullname: Marantz, Alec
  organization: Department of Linguistics, New York University, New York, NY 10003
– sequence: 3
  givenname: David
  orcidid: 0000-0003-0184-163X
  surname: Poeppel
  fullname: Poeppel, David
  organization: Department of Psychology, New York University, New York, NY 10003
– sequence: 4
  givenname: Jean-Rémi
  orcidid: 0000-0002-2121-170X
  surname: King
  fullname: King, Jean-Rémi
  organization: Meta AI, Paris 75002, France
BackLink https://www.ncbi.nlm.nih.gov/pubmed/41105708$$D View this record in MEDLINE/PubMed
BookMark eNpNkDtPwzAUhS0Eog-Y2ZBHlhT7Og97RBVQpEpdYI5s55oYJU6wm6H_nkgUiel8Ovp0hrMil2EISMgdZxvOKvE4Bp02kAMwVXGAC7LkTPGszBW7_McLskrpizGmCsmuySLnnBUVk0ty2HmMOtrWW93R5hR07y21Q-PD5xxDnEEfMdE0Itp2rvoxYosh-SFQH-ixRdpOvQ7URO3DDblyukt4e841-Xh5ft_usv3h9W37tM8sVAKyXAgUTDluTcFlbqSTjSuMkRxKB8aoRgrOQZWscQ4aXiCzWmgolNOi0gbW5OF3d4zD94TpWPc-Wew6HXCYUi2ghIorEHJW78_qZHps6jH6XsdT_XcC_ADKpmD_
ContentType Journal Article
DBID CGR
CUY
CVF
ECM
EIF
NPM
7X8
DOI 10.1073/pnas.2422097122
DatabaseName Medline
MEDLINE
MEDLINE (Ovid)
MEDLINE
MEDLINE
PubMed
MEDLINE - Academic
DatabaseTitle MEDLINE
Medline Complete
MEDLINE with Full Text
PubMed
MEDLINE (Ovid)
MEDLINE - Academic
DatabaseTitleList MEDLINE
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Sciences (General)
EISSN 1091-6490
ExternalDocumentID 41105708
Genre Journal Article
GrantInformation_xml – fundername: Abu Dhabi Institute
  grantid: G1001
– fundername: Esther A. and Joseph Klingenstein Fund (EAJK Fund)
  grantid: NA
– fundername: Whitehall Foundation (Whitehall Foundation, Inc.)
  grantid: 2024-08-043
– fundername: FrontCog
  grantid: ANR-17-EURE-0017
– fundername: NIDCD NIH HHS
  grantid: R01 DC005660
– fundername: HHS | NIH | National Institute on Deafness and Other Communication Disorders (NIDCD)
  grantid: R01DC05660
GroupedDBID ---
-DZ
-~X
.55
0R~
123
2FS
2WC
4.4
5RE
5VS
85S
AACGO
AAFWJ
AANCE
ABOCM
ABPLY
ABPPZ
ABTLG
ABZEH
ACGOD
ACIWK
ACNCT
ACPRK
AENEX
AFFNX
AFHIN
AFRAH
ALMA_UNASSIGNED_HOLDINGS
BKOMP
CGR
CS3
CUY
CVF
D0L
DIK
DU5
E3Z
EBS
ECM
EIF
F5P
FRP
GX1
H13
HH5
JLS
JSG
KQ8
L7B
LU7
N9A
NPM
N~3
O9-
OK1
PNE
PQQKQ
R.V
RHI
RNA
RNS
RPM
RXW
SJN
TAE
TN5
UKR
WH7
WOQ
X7M
XSW
Y6R
YKV
YSK
ZCA
~02
~KM
7X8
ID FETCH-LOGICAL-c2732-433e309f1cb5184b8f8df5bb8126f2bb9d83112960dff2d15e0ca3a259fa37ab2
IEDL.DBID 7X8
ISSN 1091-6490
IngestDate Sun Oct 19 01:52:27 EDT 2025
Sat Nov 01 10:37:13 EDT 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 42
Keywords language
machine learning
speech
decoding
hierarchy
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c2732-433e309f1cb5184b8f8df5bb8126f2bb9d83112960dff2d15e0ca3a259fa37ab2
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0002-9213-588X
0000-0002-2121-170X
0000-0003-4805-2910
0000-0003-0184-163X
OpenAccessLink https://www.pnas.org/doi/10.1073/pnas.2422097122
PMID 41105708
PQID 3262719238
PQPubID 23479
ParticipantIDs proquest_miscellaneous_3262719238
pubmed_primary_41105708
PublicationCentury 2000
PublicationDate 2025-Oct-21
20251021
PublicationDateYYYYMMDD 2025-10-21
PublicationDate_xml – month: 10
  year: 2025
  text: 2025-Oct-21
  day: 21
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle Proceedings of the National Academy of Sciences - PNAS
PublicationTitleAlternate Proc Natl Acad Sci U S A
PublicationYear 2025
References 38659750 - bioRxiv. 2025 Mar 03:2024.04.19.590280. doi: 10.1101/2024.04.19.590280.
References_xml – reference: 38659750 - bioRxiv. 2025 Mar 03:2024.04.19.590280. doi: 10.1101/2024.04.19.590280.
SSID ssj0009580
Score 2.4927907
Snippet Speech comprehension involves transforming an acoustic waveform into meaning. To do so, the human brain generates a hierarchy of features that converts the...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage e2422097122
SubjectTerms Adult
Brain - physiology
Comprehension - physiology
Female
Humans
Language
Magnetoencephalography
Male
Speech - physiology
Speech Perception - physiology
Young Adult
Title Hierarchical dynamic coding coordinates speech comprehension in the human brain
URI https://www.ncbi.nlm.nih.gov/pubmed/41105708
https://www.proquest.com/docview/3262719238
Volume 122
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV07T8MwELaAMrAA5VleMhIDDAEnsWNnQghRdaF0AKlb5FfUSigJDfD7uUtSwYKExBJniRSd73zf2efvI-RCCOsTYVngeWoCbqQItHJ5AFDchlpZaRxvxCbkeKym03TSbbjVXVvlck1sFmpXWtwjvwGYEUmEI-q2egtQNQpPVzsJjVXSiwHKYGDKqfpBuqtaNoI0DBKesiW1j4xvqkLX15CeIuRQQuXc3_Blk2eGW__9w22y2SFMete6RJ-s-GKH9LsYrullRzR9tUueRnO8ftyoobxS12rTU1tiOoMBytJ5gVCU1pX3dkax_XzhZ9jyXhZ0XlAAj7QR-aMGlSb2yMvw4fl-FHQCC4EF1NLclvIxS_PQGgGVnlE5TJMwBpJ-kkfGpE7FiMcS5vI8cqHwzOpYQ8WU61hqE-2TtaIs_CGhVoVQo4c25oZzKxW8O5kK4zSzIjZsQM6XRsvAgfFUQhe-_Kizb7MNyEFr-axqmTYyHqIMMVNHf_j6mGxEqM0LeSUKT0gvh_D1p2Tdfr7P68VZ4xnwHE8evwDq1sPu
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Hierarchical+dynamic+coding+coordinates+speech+comprehension+in+the+human+brain&rft.jtitle=Proceedings+of+the+National+Academy+of+Sciences+-+PNAS&rft.au=Gwilliams%2C+Laura&rft.au=Marantz%2C+Alec&rft.au=Poeppel%2C+David&rft.au=King%2C+Jean-R%C3%A9mi&rft.date=2025-10-21&rft.eissn=1091-6490&rft.volume=122&rft.issue=42&rft.spage=e2422097122&rft_id=info:doi/10.1073%2Fpnas.2422097122&rft_id=info%3Apmid%2F41105708&rft_id=info%3Apmid%2F41105708&rft.externalDocID=41105708
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1091-6490&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1091-6490&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1091-6490&client=summon