The document discusses the neuroscience of language and how our understanding has evolved over time. Early models like Broca's and Wernicke's localized language functions but did not fully capture the complexity. Modern research uses techniques like brain imaging, ERP studies, and neuropsychology to develop new models that examine both the neural structures and timing involved in integrated language processing throughout the brain.

1. The Neuroscience of
Language
From Structure to Process

2. Neuroscience and Language
• Aphasia (or dysphasia) – Language
disorder of auditory or oral speech,
writing (agraphia), or reading (alexia)
produced by injury to brain areas
specialized for these functions

3. Nature’s Experiments

4. Nature’s Experiments
• Stroke
– Status post 5 hours

5. Nature’s Experiments
• Status post 5 days.

6. Blood-Brain Barrier
• Blood is toxic to
neurons.
– Makes evolutionary
sense.
– Insulates brain from
what’s in the blood.

7. Paul Broca
• Broca (1865)
described patients
who displayed
halting, agrammatic
speech
– Content words were
well preserved
Paul Broca
1824-1880 – Function words (i.e.,
adjectives, articles)
impaired

8. Broca’s Aphasia
• Patient “Tan”
• Brain tumor in Left frontal brain region
• Broca: Lesion disrupted speech

9. Broca’s Aphasia
• Broca’s Aphasia – Damage to “motor images”
• Language comprehension skills relatively
preserved
• Typically observed in patients with damage to
left inferior prefrontal cortex

10. Broca’s Aphasia
• “Yes… ah… Monday… er… Dad and Peter
H… (patient’s name), and Dad… er…
hospital… and ah… Wednesday…
Wednesday, nine o’clock… and oh…
Thursday… ten o’clock, ah doctors… two…
an’ doctors… and er… teeth…yah
Goodglass & Geschwind, 1976

11. Carl Wernicke
• Wernicke (1874)
described patients
whose speech is
fluent, but has no
informational value

12. Wernicke’s Aphasia
• Neologisms
• Speech appears to have no
information content
• “fluent nonsense”
• Preserved function words,
impaired content words
• Comprehension impaired
• Even simple sentences not well
understood
• Associated with left temporal
lobe damage

13. Wernicke’s Aphasia
• “Well this is… mother is away here working
her work out o’here to get her better, but
when she’s looking in the other part. One
their small tile into her time here. She’s
working another time…”
Goodglass & Geschwind, 1976

14. Articulatory Auditory speech
speech “memories” “memories”
Geschwind 1979
Scientific American

15. Articulatory Auditory speech
speech “memories” “memories”
Production Comprehension
problems problems
+
empty error-filled
production

16. Broca’s aphasia Wernicke’s aphasia
Production Comprehension
problems problems
+
empty error-filled
production

17. Lichtheim
• Lichtheim (1885) – cases of patients
able to understand and produce speech
but unable to repeat words
• “Conduction aphasia”
• Hypothesis: Broca’s area (speech
production) and Wernicke’s area
(speech comprehension) intact

18. Conduction aphasia
Error-filled speech, but
good comprehension
Broca’s aphasia Wernicke’s aphasia
Production Comprehension
problems problems
+
empty error-filled
production

19. Broca’s, Wernicke’s Area, and
Connections
• Lichtheim’s (1885) and Geschwind’s (1965)
model
• Auditory input mediated by Wernicke’s area
• Motor output mediated byBroca’s area
• Regions connected by arcuate fasciculus

20. Wernicke-Lichtheim “House” Model
Concept representations
C
Widely distributed!
M A
Motor word images Auditory word images
Broca’s area Wernicke’s area

21. Lichtheim/Geschwind Model
Association Cortex
Concepts
Ventral prefrontal Posterior Temporal
cortex Motor word Auditory word Cortex
Comprehension Arcuate Comprehension
Fasciculus
Speech motor output
Auditory input

22. Lichtheim/Geschwind Model
Concepts Association Cortex
X
Ventral prefrontal Posterior Temporal
cortex Motor word Auditory word Cortex
Comprehension Arcuate Comprehension
Fasciculus
Speech motor output
Auditory input
Broca’s Aphasia

23. Lichtheim/Geschwind Model
Concepts Association Cortex
X
Ventral prefrontal Posterior Temporal
cortex Motor word Auditory word Cortex
Comprehension Arcuate Comprehension
Fasciculus
Speech motor output
Auditory input
Wernicke’s Aphasia

24. Lichtheim/Geschwind Model
Concepts Association Cortex
X
Ventral prefrontal Posterior Temporal
cortex Motor word Auditory word Cortex
Comprehension Arcuate Comprehension
Fasciculus
Speech motor output
Auditory input
Conduction Aphasia

25. More aphasias!
Aphasia type Production Comp Repetition C
1. Broca’s Non-fluent Good Poor 5 4
1 2
2. Wernicke’s Fluent, Poor Poor
paraphasic M 3 A
3. Conduction Fluent, Good Poor
paraphasic
7 8
4.Transcortical Fluent, Poor Good
sensory paraphasic
5.Transcortical Terse, Good Good
motor echolalic
6. Global Poor Poor Poor
7. Aphemia Dysarthric Good Limited only
by dysarthria
8. Pure word Normal Very poor! As Very poor
deafness if deaf!

26. Problems
• Models assumed to map directly onto
underlying brain structures
• However, patients described as
“Broca’s aphasics” had lesions in other
areas
• Inadequate psychological concepts
– e.g., “loss of motor images”

27. Broca’s Aphasia
• Damage to Broca’s area alone is not enough to produce
Broca’s aphasia
• Usually involves Broca’s area + surrounding areas
including M1 & insula.

28. Wernicke’s Aphasia
• Damage to Wernicke’s area alone is not enough to
produce Wernicke’s aphasia
• Usually involves Wernicke’s area + surrounding areas
including MTG & angular gyrus.

29. Conduction Aphasia
Common area?
• Damage to the arcuate fasciculus has not been associated
with conduction aphasia
• Usually two lesion patterns: posterior STG (wernicke’s
areas) and/or SMG

30. Transcortical Sensory Aphasia
Common area?
• Variable lesion patterns, mostly posterior to Wernicke’s
area
• Deficit tends to be transient evolving into anomic aphasia

31. Transcortical Motor Aphasia
•Damage often anterior and/or superior to Broca’s area

32. Global Aphasia
• Tend to be large “peri-Sylvian” lesions
• But smaller lesions can also cause global aphasia

33. Aphemia
• Damage to lower motor cortex

34. Pure Word Deafness
• Damage to STG bilaterally is the most common pattern,
although small left subcortical lesions have also been
documented.

35. Neuropsychology Methods
• Early neuropsychology (Broca,
Wernicke, Luria, et al.) made inferences
based largely on associations
– Identify a group of patients with fairly
homogeneous functional impairment
– fairly homogeneous physical impairment
– Identify a “syndrome”

36. Problems
• Absence of rigid quantitative
methodology of the day
• Patients not always carefully described

37. Resurgence of
Neuropsychology
in the 1960s
• Renewed interest in using
neuropsychological cases to draw inferences
about neural substrata of cognition
• More powerful inferential methods
• Emphasis on
– Single case studies
– Modern quantitative methods

38. Neuropsychology Methods
• Modern neuropsychology (e.g.,
Warrington, McCarthy, Farah)
• Study single cases or groups with
homogeneous physical impariment
• Focus on dissociations

39. Neuropsychology Methods
• Single dissociation
• Patient with lesion in brain region A
• Performs well on task A
• Performs poorly on task B
• Inference: brain region A mediates
performance on task B but not task A

40. Neuropsychology Methods
• Problems with single dissociations
• Does not rule out other interpretations
• Patient may suffer some global deficit (e.g.,
general cognitive resource) that Task A does
not require (e.g., because it is easier)
• Thus, performance differences between
Tasks A and B may reflect task difficulty, not
the function of Brain Region A

41. Neuropsychology Methods
• Double dissociation
• Patient with lesion in brain region A performs
well on Task A and poorly on Task B
• Patient with lesion in brain region B performs
well on Task B but poorly on Task A
• Inference: Brain region A mediates Task B
Brain Region B mediates Task A

42. Neuropsychology Methods
• Since subjects are “equated” for
damage, “global deficit” argument
circumvented
• Tasks may be assumed to be modular,
depending on separate neural
processes or resources

43. Brain Imaging
• We now have
techniques that
allow us to
visualize the
structure and
function of the
brain.

44. MRI
• MRI of a typical
brain.

45. PET
• Typical brain
visualized with
Positron Emission
Tomography.

46. Functional
Neuroimaging
-
time (seconds)
Deoxy
Oxy
• Functional magnetic resonance imaging

47. Functional Neuroimaging
Passive word viewing
-Activates visual areas
-Particularly Left Occipital Cortex
Passive word listening
-Activates auditory areas
-Words (not nonwords)
particularly Left Temporal Cortex
(Wernicke’s area)
• Petersen et al., 1988 PET study

48. Functional Neuroimaging
Repeating words
-Bilateral activation of motor and
Sensory face areas
-No Activation of Broca’s or
Wernicke’s
Verb Generation
-e.g., Cake  “Eat”
- Left Frontal Lobe Activation
(Broca’s Area)
• Confirms neuropsychology research

49. What Have we Learned?
• What does localization tell us?
– Where processing is occuring.
– What is connected to what.
– How many parts there are.
– What happens when one part gets
damaged.
• We’d like to know more.

50. What Would we Like to Know?
• How the parts work.
– What are the rules and representations
that are involved with language
processing?
– What is the nature of the processing that is
taking place.
• Digital/discrete/autonomous
• Analog/continuous/integrated

51. ERP studies
• Semantics
– N400
• Syntax
– LAN
– P600/SPS

52. N400 (2)

53. P600

54. Haagort et al. (2003)
Het vrouwtje veegde de vloer met een oude bezem gemaakt van twijgen (The
woman wiped the floor with an old broom made of twigs).
*Het vrouwtje veegde de vloer met een oude kliederde gemaakt van twijgen (The
woman wiped the floor with an old messed made of twigs).

55. P600 can serve as an index of
frequency of a construction
• The captain liked the crew was unhappy
• The captain heard the crew was unhappy
• The captain believed the crew was unhappy
• A P600 was strongest in the first sentence, present in
the second sentence, and not present in the third
sentence.
• Larger P600 for syntactically complex sentences
(Kaan et al., 2002).

56. Correct: De houthakker ontweek de schroef op dinsdag (The lumberjack dodged the
propellor on Tuesday).
Weak violation: *De houthakker ontweek de schroeft op dinsdag (The lumberjack
dodged the propelled on Tuesday).
Strong violation: *De houthakker ontweek de omdat op dinsdag (The lumberjack
dodged the because on Tuesday).

57. Combining methods

59. Summary
• The Wernicke-Geschwind model is
being replaced by new models.
• New models focus on both neural basis
and timing of language comprehension,
and have moved beyond simplistic
divisions of comprehension/production.