A patient with semantic dementia loses the ability to identify or describe common objects (cannot say what a camel is or draw one from memory), yet her speech production and phonological processing remain relatively intact. Which region is most likely damaged?
ABroca's area (left inferior frontal gyrus), the classical language production region
BWernicke's area (posterior superior temporal gyrus), the classical comprehension region
CThe anterior temporal lobe (temporal poles), the amodal semantic hub that integrates conceptual knowledge across modalities
DPrimary motor cortex, which controls the speech articulators
Semantic dementia preferentially damages the anterior temporal lobes (temporal poles), which serve as amodal convergence zones integrating conceptual knowledge across visual, auditory, tactile, and motor modalities. Damage here produces loss of conceptual knowledge (what things are) while leaving phonology and production relatively intact — exactly the dissociation described. Broca's and Wernicke's damage would produce different patterns: production deficits or comprehension/fluency deficits, not pure conceptual loss.
Question 2 Multiple Choice
During an fMRI study, participants read sentences like 'she kicked the ball.' Researchers using the classical Broca-Wernicke model predict that motor cortex will remain inactive, since reading is a purely symbolic, linguistic activity. What does the distributed language model predict instead?
AMotor cortex should remain entirely silent; only left temporal and frontal language areas should activate
BLeg motor cortex should partially activate as part of embodied simulation — the brain runs a partial sensorimotor simulation of the described action during comprehension
COnly visual cortex should activate to process the word 'ball'; no motor activation is expected for reading
DOnly Wernicke's area activates for action verb comprehension
The distributed/embodied language model predicts that understanding language about physical actions involves partial reactivation of the sensorimotor systems normally used to perform those actions. Reading 'she kicked the ball' should activate leg motor cortex; reading 'she twisted the doorknob' should activate hand and arm motor areas. This embodied simulation is part of how the brain extracts full meaning — language is not an isolated symbolic module but a system that recruits the brain's existing knowledge about action and perception.
Question 3 True / False
According to the classical two-area (Broca-Wernicke) model, damage exclusively to non-classical language areas like the anterior temporal lobe should produce no language deficits.
TTrue
FFalse
Answer: False
This is the misconception that distributed language network research corrects. The classical model, built from stroke lesion evidence, treats language as localized to Broca's and Wernicke's areas. But semantic dementia patients with anterior temporal damage show severe loss of conceptual knowledge despite preserved phonology and speech production. Patients with damage to other non-classical areas can show selective deficits in syntactic processing or semantic selection. The distributed network model predicts (correctly) that damage to any important node in the network will cause language deficits.
Question 4 True / False
Motor cortex partially activates when a person reads a sentence describing a physical action, even when the reader is sitting completely still.
TTrue
FFalse
Answer: True
This is a well-replicated finding in cognitive neuroscience and a key piece of evidence for embodied simulation in language comprehension. The activation is somatotopically organized — leg motor cortex for kicking sentences, hand motor cortex for grasping sentences — suggesting the activation reflects genuine simulation of the described action, not a general attention effect. This finding fundamentally challenges the view of language as an isolated, amodal symbolic system.
Question 5 Short Answer
Why does the distributed and embodied view of language change our understanding of what it means for language to be 'in the brain'?
Think about your answer, then reveal below.
Model answer: The classical view treated language as a dedicated module in specific left-hemisphere areas — language 'was' in Broca's and Wernicke's areas. The distributed view shows that meaning is computed across the whole brain's knowledge systems: the anterior temporal lobe integrates conceptual knowledge across modalities, sensorimotor cortex simulates described actions, and temporal-parietal regions handle semantic selection. Language is not a self-contained system but a coordinating process that recruits perception, action, and memory systems built for non-linguistic purposes. This means language cannot be understood in isolation from the rest of cognition.
This reframing has implications for how we understand both language acquisition (you can't learn language without building the sensorimotor and conceptual knowledge it draws on) and language disorders (deficits can arise from damage anywhere in the distributed network, not just classical areas). It also challenges the modular view of mind — language being broadly distributed suggests the boundaries between language and other cognitive systems are permeable.