Questions: Word Recognition and Lexical Access

Question 1

A reader encounters 'bank' in the sentence 'She sat by the bank and watched the water.' Eye-tracking shows a brief comprehension delay. What does this most directly suggest about lexical access?

• A. The reader did not know the word 'bank' and had to retrieve it from long-term memory
• B. Context prevented the financial institution meaning from being activated, but the river meaning took time to identify
• C. Both meanings of 'bank' were simultaneously activated, and context suppressed the financial meaning a moment later
• D. The high frequency of the financial institution meaning slowed down activation of the contextually correct meaning

Answer: (C)

Research on ambiguous words consistently shows that both meanings are activated in parallel during lexical access, regardless of context. Contextual information operates with a slight delay — it resolves the competition after the fact rather than preventing the wrong meaning from activating at all. The comprehension delay reflects this competition: the financial meaning was activated and had to be suppressed. Option B is the classic misconception — context doesn't block irrelevant meanings from activating; it selects the winner after multiple candidates have already entered the race.

Question 2

A word with many neighbors (words differing by one letter, like 'cat' → 'bat,' 'hat,' 'mat') should be recognized:

• A. Faster, because activation spreads through a densely connected network
• B. At the same speed, because neighbor words are separate entries that don't interfere
• C. More slowly, because more competitors are activated simultaneously and must be suppressed
• D. Faster, because high-frequency words like 'cat' have elevated resting activation regardless of neighborhood

Answer: (C)

High-neighborhood words activate more competitors in parallel, which slows recognition because competition-resolution must suppress more alternatives before the target wins. This is direct evidence that recognition is a competitive process, not a single-path lookup. Option A confuses spreading activation (which does occur) with speed — more spread means more interference in this case. Option D conflates two distinct phenomena: frequency raises resting activation (a real effect), but the question specifically asks about the neighborhood effect, which adds competition overhead.

Question 3

Even in a sentence context that makes one meaning of an ambiguous word obvious, both meanings are briefly activated in the mental lexicon.

• T. True
• F. False

Answer: True

Ambiguous-word studies using eye-tracking and priming consistently show that initial lexical activation is 'promiscuous' — context operates with a short delay, so the irrelevant meaning is activated before context can suppress it. This is not poor comprehension; it is the architecture of the parallel-activation system. Skilled readers resolve the competition faster, but they still activate both meanings. This finding rules out purely top-down models in which context filters activation before it begins.

Question 4

Word frequency effects in recognition demonstrate that the mental lexicon is organized like a dictionary, with common words positioned near the beginning for faster sequential lookup.

• T. True
• F. False

Answer: False

Word recognition is parallel, not sequential — the mental lexicon is not searched from a starting point. Frequency effects arise through resting activation levels: common words are encountered more often, which raises their baseline activation in the network, so they win the parallel recognition competition faster. A sequential search model might predict faster access to common words if they're 'near the front,' but this is not how the lexicon works. The activation-competition architecture explains frequency effects, neighborhood effects, and priming within a single framework that sequential models cannot.

Question 5

Why is word recognition better described as a 'flash auction' than a 'library lookup,' and what evidence supports the auction metaphor?

Model answer: Library lookup implies a sequential, one-at-a-time search that stops when the right entry is found. Word recognition instead activates many candidates simultaneously — the mental lexicon broadcasts the perceptual input and all matching representations bid competitively. The winner is determined by resting activation (frequency), contextual priming, and perceptual match. Evidence includes: neighborhood effects (more competitors = slower win), ambiguous-word delays (two meanings both bid, context must adjudicate), and semantic priming (presenting 'doctor' raises 'nurse's' bid before any conscious search occurs).

The auction metaphor captures the key features that sequential models miss: parallelism (multiple bids placed simultaneously), weighting (frequency and context adjust bid strength), and competition-resolution as a distinct stage. Poor readers may struggle not at the activation stage but at the resolution stage — multiple activated representations stay in play too long, slowing comprehension. This reframe has practical implications for reading instruction: helping struggling readers resolve lexical competition, not just recognize individual words, may be the relevant skill to target.