Psycholinguistics-Morphology
How are these words related?
1-nurse/doctor, pillar/column, tree/leaf
2-leave/leaf, burn/urn/ though/through, red/bread
3-teach/taught, speak/speaker, goat/goats, ox/oxen
How are these words related?
A-Serenity/serene, water/watering, proof/prove, read/reread
B-receive/deceive, transmit/remit
C-transmit/transmission, redeem/redemption, fall/fell electric/electricity, ten/tithe, brother/bretheren, child/children
D-teach/teacher, friend/friendly, walk/walked
E-go/went, good/best, am/is
Participate in a psycholinguistic experiment (Choose a short one)
Lexical Decision Task
prime
target
priming effect-inhibitory, facilitory
cross-modal
lag
Example of LDT test items:
tubes
govern
jubbing
bribe
leckom
tubes
trade
flup
government
kepter
allowing
leckom
tribe
flupper
swap
rejont
allow
Types of Priming
A-real word repetition priming (tubes)
B-non-word repetition priming (leckom)
C-semantic priming (trade/swap; sea/ocean; mouse/cheese)
D-orthographic priming (tribe/bribe)
E-morphological priming (govern/government; sing/sang; walk/walking)
F-non-word morphological priming (flup/flupper),
What happens if you vary orthographic/phonological overlap?
What happens if you vary semantic relatedness?
High semantic relationship create/creation
Low semantic relationship create/creature
All words
matched on ortho overlap and semantic similarity (speaker judgments)
of 1 and 3 matched.
1-high semantic relationship, morpheme in common adapter
/ adaptable
2-Low semantic relationship, historical morpheme in common part / apartment
3-high semantic relationship, no morpheme in common scream / screech
1 and 2 primed better than 3 so it's the morph in spite of varying semantics that causes priming.
semantically transparent: friendly / friend and govern / government.
semantically opaque: depart / department and author / authority.
What role does morphology play in recognizing words?
If we parse into morphemes (un+willing+ly) then polymorphemic should take more time to recognize than monomorphemic (holding word length constant).
Various
hypotheses:
1-Full listing
2-Full parsing
3-Race between parsing and access to whole word
4-Dual-route: depends on regularity of word
Evidence for full listing:
Frequency affect
reaction time
cumulative frequency is freq of 'river' plus freq of 'rivers' (freq of r-i-v-e-r)
singular frequency is freq of 'river'
plural frequency is freq of 'rivers'
|
River |
Rivers |
|
|
Windows |
Window |
|
If cumulative freq affects RT then it looks like -s is stripped off
If individual freq affect RT then it looks like each form exists in mind
Evidence
for parsing
barking