Word Association Issues:

Deaf vs. Hearing Subjects and Signed vs. Spoken Language

By Rebecca Orton

Psycholinguistics Final Paper






There are four reactions to the question of whether people think differently or not.  I don’t mean different opinions on a variety of issues.  I mean the very act of thinking itself.  My reaction would fall under the classification of being scared of the idea.  “How can we ever communicate with people who think differently?” (Lakoff, 1987, p. 306).  The second reaction is typified by the sentence “there should be a universal language so we can minimize misunderstanding and conflicts.” (Lakoff, 1987, p. 306)  Lakoff is trying to demonstrate that language itself contributes to differences in thinking.  Thinking and communicating in one language is different than thinking and communicating in another language.  The third reaction is generally a positive reaction, showing appreciation for diversity.  For example “maybe I could learn to think differently, too, maybe it would make my life richer and more interesting” (Lakoff, 1987, p. 306).  The fourth reaction is that the idea that people think differently is nonsense.  These people believe that “people are pretty much the same all around the world, there may be difference here and there, but they don’t amount to very much” (Lakoff, 1987, p. 306).


I dread the day when we discover intelligent life other than our own and find out just how different we think in comparison.  The fourth reaction may well be the predominant reaction among the human population about the human species after such a discovery.  To me, even small differences are important if the differences are so unique that only a small segment of the population has these differences, and these differences have a profound impact on the way they think and interact with the world.  Small differences at the fundamental basic level of language organization and thinking can affect increasing abstract layers of thought to the point where people behave differently to the same situation.


One fundamental and basic level of language organization is how words are organized in the mind.  Words are associated with each other in different ways.  According to Reeves, Hirsh-Pasek and Golinkoff (1998, p. 168), there are several different ways words are associated with each other.  Some words are organized together because they are opposite in meaning, for example, antonyms such as wet/dry and dark/light.  Some words have associated words that complete a pair commonly used or understood as part of real life, for example, salt/pepper and king/queen.  Words are most likely to be associated with other words within the same grammatical class, for example, noun with noun (chair/table), verb with verb (run/jump), and adjective with adjective (dark/light).  Some words are associated with other words because they are a subordinate or superordinate  class of those other words, for example, apple/fruit and chair/furniture.  Last and most importantly, words are most likely to be associated by their meanings. 


Incidentally, perceptual similarity is not considered to be an organizational basis for word associations for English.  Long and thin shapes are not categorized together, for example, needle/thread or needle/sew rather than needle/nail or needle/poker (for the fireplace), both have long and thin shapes.  I believe that needle/pin does show some perceptual similarity.  In addition, the classifier system in American Sign Language does seem to have some basis on perceptual similarity.  For example, the number one handshape classifier is used to represent long and thin shapes, most commonly a human.


According to Reeves, Hirsh-Pasek and Golinkoff (1998, p. 168), “There is no reason in principle why subjects in word associations tasks would not choose similar sounding words (as in wheedle for needle), but they don’t.”


However, I found, much to my surprise, that I have a phonological as well as a semantically based organization of words that manifested itself when I took the following word association test:  wet/bet, swift/stiff, petal/metal, apple/orange, shoot/hoot, needle/poodle, salt/pepper, queen/king, chair/sofa, and dark/light.  I did use semantically based word pairs (salt/pepper and king/queen).  I used words that were the members of the same class (apple/orange, chair/sofa).  However, most of my word associations were not based on meaning, but on phonological similarity (wet/bet, swift/stiff, petal/metal, shoot/hoot, and needle/poodle).  My whole psycholinguistics class noticed that my responses were different from theirs.  I was the only deaf person in the class that took the word association test.  The rest of the class that took the test was hearing.  The most obvious reason that could be attributed to my fundamental difference was my deafness.


I reasoned that since I am deaf, I needed to have a word organization based on phonological similarity as well as semantic similarity.  I could process words that I heard through my hearing aid, and match them up with the visual lip-reading input that helped to phonologically make the words distinctive enough to understand them.  Most speech sounds are alike when heard with only one-fifth of the normal frequency range of the human ear, which is what I have.  In addition, lip-reading adds about thirty percent more clarity.  The necessity of perceiving words from two ineffective inputs would motivate a phonological-semantically based system of word organization in order to increase the effectiveness of language processing to the point where I could understand what was being said, and thus interact with the world out there.


The idea that this system of phonological-semantically based word organization that I have developed in response to the need for survival in the hearing world is the very system that makes me so unique, so different, is scary.  In hopes of relieving my fears, I decided to investigate the possibility that other deaf people have the same kind of phonological-semantic lexical organization.  I also included hearing people in my research as a comparison group.





I selected three pools of subjects to cross-validate my data.  One pool was strictly a hearing group of six subjects with hand-written responses to a test that was already given in my psycholinguistics class prior to my collecting their tests.  The second pool was mixed group of 65 hearing and deaf subjects with typed responses to an email survey.  The third pool of six subjects signed their responses back to me to a revised test that I signed to them.  I wrote down the corresponding English glosses for their signed responses, each sign at a time.


The six subjects in my psycholinguistics class were not familiar with sign language in general except for one subject who knew a high school teacher with a daughter that is a student at Gallaudet University.  Their first sign language exposure was my interpreter for the class.  However, it was clear that the subjects were not strictly monolingual English speakers. 


The subjects of the email survey were from diverse backgrounds.  I received responses from 65 subjects, but not all responded with data that I could use.  36 subjects indicated that they were deaf, 17 responded that they were hard of hearing.  Some subjects indicated both deaf and hard of hearing, and both responses were included in the counts of deaf and hard of hearing totals.  17 subjects were hearing.  These totals are not meant to add up to the 65 total.   The numbers in this paper are only meant to determine which populations are represented.


54 subjects were native US citizens, two were immigrants, and one subject indicated dual citizenship in USA and Canada.  Other countries represented by the subjects were Canada – 3, Malaysia – 1, Austria – 1, Denmark – 1, and Switzerland – 1.


49 subjects spoke English fluently and 62 wrote English fluently.  39 subjects signed ASL (American Sign Language) fluently.  31 fluently signed PSE (Pidgin Signed English), otherwise known as contact sign or sign supported speech.  One subject signed SEE, Signing Exact English, which is a manually coded English system in visual form. 


Other languages represented by subjects in the email sample are Spanish – 18, French – 13, German – 9, Hebrew – 3, Russian – 3, Chinese – 2, Danish – 2, Gestuno (International Sign Language) – 2, gestures – 2, Italian – 2, Italian Sign Language (LIS) – 2, Malay – 2, New Zealand Sign Language – 2, Norwegian – 2, Ancient Greek – 1, Arabic – 1, Austrian Sign Language – 1, Chinese Sign Language – 1, Costa Rica Sign Language (LESCO) – 1, Danish Sign Language – 1, Danish Tactile Sign Language (for deaf-blind people) – 1, French Sign Language – 1, German Sign Language – 1, Hebrew Sign Language – 1, Japanese – 1, Latin – 1, Malay Sign Language – 1, Norwegian Sign Language – 1, Romanian Sign Language – 1, Swedish – 1, Swedish Sign Language – 1, Swiss – 1, Vietnamese – 1, and Yiddish – 1.


As shown above, email survey results are from both hearing and deaf subjects, monolingual, bilingual, and multilingual Americans and non-Americans alike.  However, only one subject, a deaf woman from Austria, shows this same heavily influenced phonological-semantically based system of word association that I have.  Her word results were:  wet/wet, swift/sweet, petal/penis, apple/apricot, shoot/shit, needle/pinpoint, salt/pepper, queen/king, chair/sofa, and dark/light. Clearly, the reason for this difference is not deafness in general.  I needed to analyze in more depth additional background data of this Austrian subject to discover whether or not there are other reasons for this difference.


The six subjects for the signed ASL association test were from the same pool of subjects as the email survey except for one deaf man fluent in ASL.  Using the same subjects from the email pool for the signed test pool might influence the signed test results because the subjects were already aware that I had given them a similar test before.  But because the signed test was in a different modality (signed, not written) and I was looking for phonological influences in that modality, giving a similar association test to the same subjects was not a concern.



I collected six written word association tests from my psycholinguistics class.   The test was typed with two columns and the responses were hand-written.  The written test is included at the end of the paper in Appendix A.


I sent out an email survey with the same test words organized in one column.    The email survey is included at the end of the paper in Appendix B.  English responses were in lower caps and ASL responses were in upper caps.  For the most part, the email subjects were able to follow directions, but for some, the data needed to be interpreted more flexibly.


I designed a similar test with American Sign Language (ASL) as the language medium.  I signed a similar list of words to six subjects fluent in ASL.  The subjects signed back to me and I wrote down the English glosses that represented their signs.  Here I capitalize the English glosses to represent the ASL signs, as is the custom within the linguistics program at Gallaudet University.  The test list is:  WET, FAST, LAW, APPLE, SHOOT, SEW, SALT, QUEEN, CHAIR, and DARK.  You can find this test list in Appendix C as well. 


As I was getting email responses and analyzing them on the fly, I also sent an email to the Austrian subject asking for more background information and she responded before I finished collecting all my data.





One of the first things I thought of was to ask this Austrian subject some pertinent questions to see if I could track down what was common to both of us and yet unique compared to other people.  I was able to do a preliminary analysis based on her email reply because her responses were very revealing.


The first thing I noticed in her reply was that she had slightly better hearing in her left ear than she did in her right, about 10 dB better.  She is still a deaf person, but has some residual hearing.  I can hear in my left ear better than in my right ear.  I immediately thought of the right ear advantage as mentioned by Dingwall (1998, pp. 79-80).  The right ear advantage occurs when speech stimuli is presented at the same time to both ears.  The left hemisphere gets inputs from the right ear directly and from the left ear indirectly through the commissures (neural tissue connecting the two hemispheres).  With me, my left hemisphere is getting one or two inputs, one directly from my left ear and, supposedly, one indirectly from my left ear through the commissures. I hear monaurally.  In binaural hearing, the neural pathways from the ears to the hemispheres on the same sides are suppressed, but not on the contralateral sides.  The contralateral neural pathways from the ears to the opposite hemispheres are never suppressed.  In monaural hearing, the neural pathways from the ear to the hemisphere on the same side are not suppressed (Dingwall, 1998, pp. 79-80).  It is possible that a difference in hearing in the better left ear for the Austrian subject could have had an effect on which neural inputs her left hemisphere has.  She has two hearing aids, but her left ear is stronger.  The suppression effects of binaural hearing may not be as strong or applicable with a difference in hearing between both ears.  The Austrian subject could, in effect, be hearing with the neural inputs of monaural hearing, like me.  However, a small difference such as 10 dB makes it seem unlikely.


The left hemisphere is where most of the linguistic processing functions are normally located (Dingwall, 1998, pp. 81-83).  The problem with assuming that linguistics functions are located in the left hemisphere is that this may not be true of the Austrian subject and me.  We show tendencies to be more ambidextrous.  For example, I can write with my left hand even though I normally write with my right.  The Austrian subject prefers to sign with her left hand as the dominant hand even though she can sign with her right hand as the dominant hand.  This suggests a more bilateralization of language functions within our brains. 


Being deaf from birth may affect lateralization of language functions as well.  Both the Austrian subject and I were born deaf.  According to Corina’s fMRI study (1998, pp. 42-45), for written English stimuli, hearing native signers and hearing non-signers showed the same pattern of left-hemisphere activation.  Deaf native signers had fewer areas of left hemisphere activation for written English stimuli.  For ASL stimuli, hearing native signers and deaf native signers showed the same bilateral pattern of activation, all the left areas associated with language, plus many right areas.  The hearing non-signers had no significant activation.  Deaf native signers had more right areas activated than hearing native signers, which may be due to their deafness.


The next thing I thought of was that the right hemisphere was where creativity and prosody originates from.  According to Dingwall (1998, p. 80), musical stimuli and visual-spatial stimuli are processed better with the right hemisphere than with the left.  Being a poet or an amateur musician would be suggestive of a more bilateral distribution of language functions.  However, professional musicians process music through the left hemisphere rather than the right because of their technical knowledge and analytical approach to music (Ortega, 2000, personal communication). I like to write poetry and sing.  The Austrian subject indicated that she liked to write poetry when she was inspired.  She believes that she is creative when she wants to be as well.  She can also talk.


Another subject that I found both ASL and English phonological influence on his email survey results is a famous deaf man that writes books about new and creative words describing aspects of deaf culture.  His word associations were in ASL but reflected English associations as well:  wet/WATER, swift/FAST, petal/PEDAL, apple/FRUIT, shoot/SHUCKS, needle/OUCH, salt/PEPPER, queen/KING, chair/SIT, dark/BRIGHT.


Having a creative bent may indeed explain why my word associations were so heavily influenced by phonology instead of being purely semantic.  Fortunately, there is a more plausible explanation that was evident after I had collected all my data and analyzed the results I received in more depth.





Data Validation


I looked at what kinds of semantic associations result from this new ASL association test to see if they are similar to the kinds of word associations found from the first test as discussed by Reeves, Hirsh-Pasek and Golinkoff (1998, p. 168).  I found that the salt/pepper association pair was common among the results of all three methods of testing.  The written test had five entries of the salt/pepper word pair.  The email survey had 24 English salt/pepper entries and four ASL SALT/PEPPER entries.  The signed data had two entries of SALT/PEPPER.  This similarity of data across the three methods is assurance that the data is indeed valid and analyzable.


I found basic level, subordinate level and superordinate level categorical associations among the results of all three methods except for the written test in the basic/superordinate category.  See Table 1 starting on next page.  I also included a count of how many entries were found if there were more than one entry for an association pair.  These associations are not the same associations like with salt and pepper mentioned above, although there are many association pairs that are common between English and ASL.  This hypernymic/hyponymic/same membership grouping is further reassurance that the data is valid and analyzable across the three methods.



Table 1.  Semantic associations by hypernymity/hyponymity/same membership


Basic/superordinate categorical pairs are:


Signed Data:












SEW/C-R-A-F-T (see note about P-E-A-C-H)





Email Data:

apple/fruit (English, 10 entries found)

apple/FRUIT (ASL, 3 entries found)




























Written Data:

None found applicable

Basic/subordinate categorical pairs are:


Signed Data:







CHAIR/HIGH CHAIR (also included as phonological-semantic pair)









Email Data:









chair/throne (this entry was semantically influenced by previous word queen in the list)




















Written Data:




chair/throne (Again, this entry was semantically influenced by the previous word queen in the list)

Same categorical pairs are:


Signed Data:

APPLE/P-E-A-C-H (fingerspelled semantic association only, no phonetic association possible within ASL)


CHAIR/D-E-S-K (see note about P-E-A-C-H above)










Email Data:

apple/orange (English, 3 entries found)

apple/ORANGE (ASL)

apple/peach (English, 2 entries found)

apple/pear (English, 2 entries found)


chair/table (English, 9 entries found)

chair/desk (English)

chair/couch (English)

chair/sofa (English)


needle/pin (English, 2 entries found.  This grouping has a visual similarity as well as a categorical similarity, both needles and pins are long and thin objects. Perhaps an image schema of long and thin objects was at work here in this word association.)


petal/leaf (English)

petal/LEAF (ASL)


petal/stem (English)


Written Data:

needle/pin (see note above about same entry)


chair/table (2 entries found)





I also found opposite association pairs and complementary association pairs among the results of all three methods as shown in Tables 2 and 3, respectively.  Again, these groupings are more evidence that the data is valid and analyzable.


Table 2.  Opposite pair associations


Signed Data:



LAW/O-U-T LAW (This is also a phonological-semantically opposite pair.  The sign LAW is common to both, with O-U-T fingerspelled.  The sign pair is semantically related to each other.  For example, an outlaw breaks the law.)

Email Data:
















dark/BRIGHT (ASL, 2 entries found, this entry is also an phonological-semantically opposite pair.  Only an orientation change within the hands occurs.  Semantically, dark is the opposite of bright.)

dark/bright (English),

dark/light (English, 20 entries found),

dark/LIGHT (ASL 3 entries found)


swift/slow (English, 4 entries found)


wet/dry (English, 9 entries found)

wet/DRY (ASL)

Written Data:


























dark/light (2 entries found)



Table 3.  Complementary pair associations


Signed Data:

























SHOOT/KILL (two entries found)











Email Data:

queen/king (English, 23 entries found),

queen/KING (ASL, 2 entries found),



dark/night (English, 9 entries found),

dark/NIGHT (ASL)


needle/thread (English, 10 entries found)

needle/THREAD (ASL, 2 entries found)


petal/flower (English, 27 entries found)

petal/FLOWER (ASL, 6 entries found)

petal/rose (English, 3 entries found)

petal/ROSE (ASL, 2 entries found)




shoot/gun (English, 19 entries found)

shoot/GUN (ASL, 7 entries found)

Written Data:

queen/king (4 entries found)














petal/flower (3 entries found)










shoot/gun (4 entries found)







Phonological-Influenced Data


Now that I am convinced that the data is valid and worth analyzing, I looked for any phonologically-semantically based associations regardless of modality.  Please see Table 4 below.  Again, as before, I included a count of how many entries were found if there were more than one entry for an association pair.  Please note that all six ASL subjects contributed to the data in the Signed Data section in Table 4.  As is evident in Table 4, a lot of English and ASL subjects contributed to the data in the Email Data section.  Between four and five English subjects contributed to the data in the Written Data section in Table 4.


Table 4.  Phonological-Semantic pair associations


Signed Data:










CHAIR/SIT (two entries found, movement change














CHAIR/HIGH CHAIR (phonetically, one sign in common, also within basic category/subordinate category)





















DARK/SAD (movement and handshape change only)


LAW/RULES (handshape change only)

















































SHOOT/HUNTING (handshape change only)

SHOOT/GUN (two entries found, a change from a

two-handed to one-handed sign)






































































WET/SEX (This entry is a phonetic-semantic pair

based only on English rhyme, not a signed

phonetic-semantic pair even though this entry

is signed ASL data, the associations themselves

appear to be based on English phonology.)

Email Data:

apple/apricot (English, both words have initial syllables that sound the same phonetically.  The concept of apple is in the same basic level category as the concept of apricot semantically.)


chair/SIT (ASL, 10 entries found, only a movement change would be involved phonetically if chair was associated with the ASL sign CHAIR first and then associated with the ASL sign SIT.  The concepts of chair and sit are semantically similar.  There was one entry where both the English word sit and the ASL sign CHAIR were entered.)









chair/SWING (ASL, only a movement change would be involved phonetically if chair was associated with the ASL sign CHAIR first and then associated with the ASL sign SWING.  The concept of swing is within a subcategory of the concept of chair, which is a basic level item.)


dark/dim (English, both start with the same letter, and is semantically similar as well)








queen/KING (ASL, 2 entries found, again, only a handshape change from a Q-handshape to a K-handshape would be involved phonetically if queen was associated with the ASL sign QUEEN first and then associated with the ASL sign KING.  The concepts of queen and king are semantically similar.  There was one entry where both the English word king and the ASL sign KING were entered.)


queen/ROYALITY (ASL, R-handshape, see note above about queen/KING)


queen/PRINCE (ASL, P-handshape, see note above about queen/KING)


queen/PRINCESS (ASL, only a handshape change to a P-handshape and an additional movement to another similar location on the chest would be involved phonetically if queen was associated with the ASL sign QUEEN first and then associated with the ASL sign PRINCESS.  The concepts of queen and princess are semantically similar.) 


salt/spice (English, both words phonetically start with the same s sound and are semantically similar as well.)




shoot/GUN (ASL, 7 entries found, again, only a weak hand drop and a movement change would be involved phonetically if the English word shoot was associated with the ASL sign SHOOT first and then associated with the ASL sign GUN.  The concepts of shoot and gun are semantically similar.  There was one entry where both the English word gun and the ASL sign GUN were entered.)


shoot/shot (English, this entry is a very good example of an English phonological-semantic association, with only one phonetic change in the vowel, similar to the more common ASL phonological-semantic entries found above.

There may also be an orthographic-semantic

association as well since the letter o was dropped in the second word.)


shoot/shit (English slang, 2 entries found, both start with the same sh phonetic sound and can be a semantic expression of disappointment.  There is also an entry similar to this one, shoot/SHIT in ASL. This entry could be a case of first the English word phonetic-semantic associations of shoot/shit and then the semantic association of the ASL sign occurring.)


shoot/shucks (English slang, again, this entry is similar to the entry shoot/shit above as well)


shoot/straight (English, This entry is more heavily influenced by phonology than by semantics.  The beginning and ending of the word pair are similar.  There is a semantic relatedness as well.  You shoot straight, for example, in order to hit a target.)


swift/quick (English, 4 entries found, both have the same phonetically rhyming vowel sound and mean the same thing semantically.)














wet/water (English, 14 entries found, this word pair is closely related both phonetically and semantically.  Both initial syllables start with w and end with t, and water is certainly wet.)


wet/willy (English, this word pair is not as closely related as wet/water above, but does start phonetically with w, and semantically means putting a wet finger in someone's ear as a gag.)


wet/STICKY (ASL, only a handshape change would be involved phonetically if the English word wet was associated with the ASL sign WET first and then associated with the ASL sign STICKY.  The concepts of wet and sticky are semantically similar.)

Written Data:












































































































































































swift/quick (see note  for same entry)


















wet/water (4 entries found, see note for

same entry)





























I looked for phonological-orthographical association pairs and found them only within the email data.  See Table 5 below.  Three English subjects contributed to this data in Table 5 below.


Table 5.  Phonological-Orthographical pair associations


Email Data:


swift/sweet (English, both words start with sw and end in t phonetically and orthographically.  There is no semantic similarity between the two words.)


petal/penis (English, both words start with pe orthographically.  The p is phonetically the same sound, but not the initial vowel. There is no semantic similarity between the two words.)


petal/pedal (English, both words sound the same except for the voicing feature.  The t is voiceless and the d is voiced.  Both words look the same orthographically, except for the middle letter.  There is no semantic similarity

between the two words.)


petal/metal (English, both words sound/look the same except for the first sound/letter.  There is no semantic similarity between the two words.)


I also found within the email data cross modal phonological-semantic associations.  See Table 6 below.  Please note that five different ASL subjects contributed to this data in Table 6 below.


Table 6.  Cross-Modal Phonological-Semantic pair associations:


Email Data:


chair/SALT (ASL, the English word chair could be associated with the ASL sign CHAIR first, and then associated with the ASL sign SALT.  The only phonetic change would be in movements of the signs.  There is no semantic similarity between the concept of chair and salt.)


salt/CHAIR (ASL, The English word salt could be associated with the ASL sign SALT first, and then associated with the ASL sign CHAIR.  The only

phonetic change would be in movements of the signs.  There is no semantic similarity between the concept of salt and chair.)


petal/PEDAL (ASL, The English word petal could be associated with the English word petal phonologically, and then associated with the ASL sign PEDAL.  There is no semantic similarity between the concept of petal and pedal.)


wet/AND (ASL, the English word wet could be associated with the ASL sign WET first, and then associated with the ASL sign AND.  Phonetically, WET and AND are similar.  Changes in orientation, movement, and from a two handed to a one handed sign would be involved.)


wet/SOFT (ASL, the English word wet could be associated with the ASL sign WET first, and then associated with the ASL sign SOFT.  Phonetically, only a movement change would be necessary.  There is no semantic similarity between the concepts of wet and soft.)



Phonologically or Semantically Influenced Data Based on Word Order


Finally, I also looked for phonological or semantic influence on associations that may appear because of the way the word association tests are ordered.  There may be either semantic or phonological influence from prior or anticipated words or signs.  However, I did not find any in my signed data.  Perhaps because I paused the test after each ASL sign response in order to write down the English gloss, the results were affected.  If a camera were used to record the ASL responses on videotape, there would be no need to write down the English glosses.  The videotape itself could be used for transcription later into English glosses.  The test timing would also be the same as that of the written test and email survey.  As was mentioned previously in our psycholinguistics class, I found semantically influenced word associations in the written data based on word order.  I also found both semantically influenced and phonologically influenced associations in the email data based on word order.  Three English subjects contributed data to the Email Data in Table 7.  As is evident, only one English subject contributed to the data in the Written Data section in Table 7.


Table 7.  Word-Order-Influenced pair associations:


Email Data:


chair/back then dark/black (English, phonologically-influenced)


chair/brown then dark/wood  (English, semantically-influenced)


queen/… then chair/throne (English, semantically-influenced)


swift/pedal then petal/flower (English, phonologically-influenced)

Written Data:








queen/… then chair/throne (English, semantically-influenced)







In the case of the famous deaf man, his associations showed strong evidence for cross-modal transfer or in other words, intersensory integration as mentioned by Wolf, Vellutino and Gleason (1998, p. 418).  That is, if connecting spoken and written representations of the same concept is one type of cross-modal transfer, then so is connecting written and signed representations of the same concept. 


Because of the evidence I have presented here, I am inclined to believe Fromkin and Ratner (1998, p. 323) when they say that the lexicon is organized both semantically and phonologically, not strictly semantically as Reeves, Hirsh-Pasek and Golinkoff (1998, p. 168) seem to suggest.  Fromkin and Ratner based their conclusions on word substitution errors and word blends.  The “words involved are semantically or phonologically similar or both” (Fromkin & Ratner, 1998, p. 323).  Their table listing different kinds of errors is shown below in Table 8.


Table 8.  English Semantic/Phonological Errors


1.  That’s a horse of another color - … a horse of another race (semantic substitution)

2.  Too many irons in the fire – too many irons in the smoke (semantic substitution)

3.  White Anglo-Saxon Protestant - … prostitute (phonological substitution)

4.  Grab/reach – greech (semantic blend)

5.  Gin and tonic – gin and topic (phonological)

6.  Arrested and prosecuted – arrested and persecuted (phonological/semantic)

7.  At 4:30 we’re adjourning the meeting – we’re adjoining the meeting (phonological)

8.  stiffer/tougher – stougher or stuffer (semantic/phonological blend)

9.  Edited/annotated – editated (semantic blend)



The same phenomena can be found in ASL as well.  I have experienced it myself.  With the ASL signs FRUSTRATED and WRONG, I mistakenly switched the hand orientation between the two signs.  With SMALL WORLD, I mistakenly switched the handshapes.  There are some found in the research literature as well.  Fromkin and Ratner found one with SICK and BORED with the location switched (1998, pp. 316-317).  Poizner et al’s (1987, p. 199) paraphasic subject KL substituted an inappropriate W handshape for a K handshape in CAREFUL, and substituted an inappropriate up and down movement for a circling movement in ENJOY.  Corina’s (1998, pp. 38-39) subject WL switched the inappropriate Y handshape for the 1 handshape in TOOTHBRUSH, and switched the inappropriate A handshape for the H handshape in SCREWDRIVER.  WL also switched the inappropriate Y handshape for the 5 handshape in FINE.  According to Corina (1998, p. 38), the most frequent errors are in handshape substitution, but substitution errors were found in all four major parameters, handshape, location, movement, and hand orientation.  See Table 9 below for a recap of these ASL phonological errors below.



Table 9.  ASL Phonological Errors


1.  FRUSTRATED/WRONG (hand orientation switch)

2.  SMALL/WORLD (handshape switch)

3.  SICK/BORED (location switch)

4.  CAREFUL (*W handshape substitution for K handshape)

5.  ENJOY (*up and down movement substitution for circling movement)

6.  TOOTHBRUSH (*Y handshape substitution for 1 handshape)

7.  SCREWDRIVER (*A handshape substitution for H handshape)

8.  FINE (*Y handshape substitution for 5 handshape)



According to Fromkin and Ratner, “the choice of inappropriate lexical items may occur because synonyms, antonyms, and similar sounding words are stored in close proximity to a given target word, and thus may be retrieved in error” (1998, p. 323).


Word associations seem to work the same way, although not meant to be in error, of course.  In retrospect, I would expect such phonological influence to show up in English and ASL.  For example, in English, we have words like glimmer, glitter, gleam, etc. that are semantically and phonologically similar.  Another example is twirl, whirl, swirl, etc.  In ASL, there are groupings of signs that operate on the same iconic principle.  For example, the ASL signs associated with thinking are usually located at the temple:  THINK, KNOW, WONDER, etc.  Another example is the ASL signs signifying some sort of group: CLASS, FAMILY, TEAM, etc. This kind of iconicity is evident in the word association data presented here as well. 


It is a relief to find that my mind is not so different as I initially thought.  That being bilingual in ASL and English as well as having a creative inclination may well explain why my test results were so heavily influenced by phonology.  Perhaps that day I took the word association test, I was feeling particularly inspired.  I do know that when I took another word association test for my  classmate in psycholinguistics class, the phonological influence was not as heavy.  The second time, I was also aware of the purpose of the word association test as well, which probably affected the test results.  So, my fears of being too unique have been put to rest.






Corina, D.P., (1998).  Studies of Neural Processing in Deaf Signers: Toward a Neurocognitive Model of Language Processing in the Deaf.  In Journal of Deaf Studies and Deaf Education (Vol. 3, #1, pp. 35-48).  Seattle, WA: Oxford University Press.


Dingwall, W.O., (1998). The Biological Bases of Human Communicative Behavior.  In J.B. Gleason & N.B. Ratner (Eds.), Psycholinguistics (pp. 51-105). Fort Worth, TX: Harcourt Brace College Publishers.


Fromkin, V.A., Ratner, N.B., (1998). Speech Production.  In J.B. Gleason & N.B. Ratner (Eds.), Psycholinguistics (pp. 309-346). Fort Worth, TX: Harcourt Brace College Publishers.


Lakoff, G., (1987).  Women, Fire and Dangerous Things  What Categories Reveal about the Mind.  Chicago, IL: The University of Chicago Press.


Poizner, H., Klima, E.S., & Bellugi, U. (1987).  What the Hands Reveal about the Brain (Ch. 8, pp. 193-212). Cambridge: MIT Press.


Reeves, L.M., Hirsh-Pasek, K., & Golinkoff, R., (1998). Words and Meanings: From Primitives To Complex Organization.  In J.B. Gleason & N.B. Ratner (Eds.), Psycholinguistics (pp. 157-226). Fort Worth, TX: Harcourt Brace College Publishers.


Wolf, M., Vellutino, F., & Gleason, J.B. (1998). A Psycholinguistic Account of Reading.  In J.B. Gleason & N.B. Ratner (Eds.), Psycholinguistics (pp. 408-451). Fort Worth, TX: Harcourt Brace College Publishers.



Appendix A: Written Test


As fast as you can, write down the first word that comes to your mind after each word:


    wet--                              needle--

    swift--                            salt--

    petal--                            queen--

    apple--                            chair--

    shoot--                            dark--


Appendix B: Email Survey


Hi Everyone,

I'm doing some research for a paper for my psycholinguistics class at Georgetown University (I'm still a Gallaudet student until I graduate this May).  I would like you to fill out this quick quiz and send it back to me.  The deadline is March 22 for gathering this data.  Could you also pass this email along to your distribution lists if appropriate.  Thanks! 


Are you

deaf? (y/n) -

hard of hearing? (y/n) -

other? (please explain) -


Are you a

native USA citizen? (y/n) -

USA immigrant? (y/n) -

citizen of other country?

(please explain) -


Do you

speak English fluently? (y/n) -

write English fluently? (y/n) -

Other? (please explain) -


Do you

sign American Sign Language fluently? (y/n) -

sign PSE fluently? (y/n) -

(PSE = contact sign or sign supported speech)

other? (please explain) -


Do you know other languages (signed or spoken)?

(please explain) -


As fast as you can, write down the first English word (or English gloss in CAPS for an ASL sign) that comes to your mind after each word:


wet -

swift -

petal -

apple -

shoot -

needle -

salt -

queen -

chair -

dark -


That's it!  Thanks for your help! 


If I find that I need face to face research, I may contact some of you who are local to my area to meet with me, but other than that, there's no further commitment.


I will post my paper when it is finished on my web page at http://www.geocities.ws/rebaorton  Again, thanks for your help!




Appendix C: Signed Test


English glosses are listed below for the corresponding ASL signs: