Journal of Undergraduate Research
Volume 5, Issue 1 - October 2003

The Effects of Phonological Priming and Word Production on Tip-of-the-Tongue Resolution

Stacy Eitel

ABSTRACT

The effect of phonological priming on tip-of-the-tongue (TOT) resolution was assessed by presenting participants with lists of words to read, either aloud or silently, while experiencing a TOT state. Without the participants’ awareness, these lists contained 2 prime words that both began with the same first syllable or first phoneme as the target word, intermixed among 8 unrelated words. Reading first syllable primes significantly resolved TOTs more than first phoneme primes or unrelated words. Furthermore, reading first syllable primes silently produced greater resolution than reading aloud. This research supports the Transmission Deficit Hypothesis that TOTs are caused by weakened phonological connections.

INTRODUCTION

The TOT state has been recognized early on as an exciting, albeit frustrating, experience. William James eloquently described the TOT state in 1893 as a state of consciousness consisting of a gap “beckoning us in a given direction, making us at moments tingle with the sense of our closeness and then letting us sink back without the longed-for term.” (Schwartz, 2002, p. 4).

Two leading theories regarding the etiology of TOTs have emerged: the blocking hypothesis and the Transmission Deficit Hypothesis. The blocking hypothesis (Jones, 1989, as cited in Schwartz, 2002) posits that TOTs occur when a related word receives more activation than the target word, thus inhibiting it. This theory predicts presenting phonological primes will decrease TOT resolution because primes will “block” retrieval of the target word.

The Transmission Deficit Hypothesis (TDH) states that connections between a word’s lexical and phonological nodes weaken if they have not been activated recently, causing word retrieval failures (Burke, MacKay, Worthley, & Wade, 1991). Figure 1 diagrams these nodes, and the darkened lines illustrate how saying pellet activates the /el/ sound shared by velcro. This theory predicts that phonological primes should increase TOT resolution by strengthening phonological connections to words.

Figure 1. Semantic and phonological nodes representing the prime word pellet and the target word velcro.
Figure 1. Semantic and phonological nodes representing the prime word pellet and the target word velcro. Many nodes necessary for producing these words have been omitted for simplification. Note that the representations for the two words overlap only at the phonological level. Phonological connections strengthened by saying pellet have been darkened. TOT = tip-of-the-tongue state. (figure from James & Burke, 2000).


James and Burke (2000) studied phonological priming of TOT states. After presenting a question, they primed participants while they were experiencing a TOT by presenting words that cumulatively contained all of the syllables of the missing word (e.g., TOT word= abdicate, primes= abstract, indigent, truncate). They found that primes increased word retrieval relative to an unrelated list of words. However, primes shared multiple phonological features with the TOT word (i.e., same first, middle, and last syllable), making it impossible to isolate which phonological feature(s) was the most effective or even essential for TOT resolution.

The present study isolated phonological features of TOT words and assessed their effect on resolution of TOTs. Two specific pieces of phonology were used: the first phoneme or first syllable of the target word. The experiment also examined word production's role in priming by having participants either read primes silently or aloud.

Based on the TDH, I hypothesize that phonological primes will strengthen the weak connections that theoretically exist between phonological and lexical nodes, aiding TOT resolution. I expect to find that first syllable primes aid resolution. I also predict that first phonemes will show positive priming effects. TDH also predicts that both modes of production should facilitate TOT retrieval, since both utilize the same phonological connections.

METHOD

Participants

Participants were 84 undergraduate, English speakers (19 males and 65 females, aged 18-23 years, M = 18.62 years, SD = 0.89), enrolled in introductory or cognitive psychology courses. They were compensated with course- or extra-credit.

Materials

The 80 general-knowledge questions used in the experiment were chosen for their tendency to elicit TOT states in previous studies (Abrams & White, 2001; White & Abrams, 2002). The word lists used to prime participants in the experimental conditions contained two primes that were phonologically related to the answer (i.e., target word) and eight filler words that were unrelated to the target. The prime words in the phoneme condition shared their first phoneme with the target, while the first syllable condition primes shared first syllable phonology with the target. The unrelated (i.e., control) condition contained the same eight unrelated filler words as the first phoneme and first syllable conditions, but two unrelated words replaced the primes from the experimental conditions (see Table 1 for examples). Primes varied in position with half of the primes in positions two and eight of the ten words and half in positions three and nine.

A multiple choice recognition test was given after all questions were presented to assess the accuracy of participants’ TOT responses. If they were experiencing a TOT state for the intended target, they should select the correct answer from the four choices. Participants also completed a Nelson-Denny Vocabulary Measure, which consisted of 25 multiple-choice vocabulary questions.


Table 1.
Examples of the Three Conditions of Stimuli for a Target Word
Question Target 1st Syllable 1st Phoneme Unrelated
What do you call a person who makes maps? cartographer cardiogram
carbohydrate		 customary
combination		 occupancy
molecular
What are the front cutting teeth called? incisors innocence
indicate

immortal
illusion

 turpentine
personal
What do you call false testimony under oath? perjury perpetrate
permafrost		 pioneer
pollution		 survivor
derision
What is the proper name for a 'tidal wave'? tsunami superscript
suicide

structura
specialize

 honeymoon
loneliest

Procedure


Figure 2 depicts a flowchart of the procedure from the experimenter’s perspective. Participants were told that we were collecting data on two tasks. The first task presented participants with 80 questions on a computer screen, and they were asked whether they knew the answer, did not know the answer, or were experiencing a TOT state. Participants were told “you experience a TOT every time you are unable to retrieve a word that you are certain you know". The second task involved collecting data on the pronounceability of certain words (in reality, this task was an attempt to disguise the relationship between the primes and target word). After each question, participants read a list of words and were asked to verbally rate how easy or difficult a word was to pronounce on a scale of one (easy) to seven (difficult). Following don’t know or TOT responses, the ten words contained either two first syllable primes, two first phoneme primes, or completely unrelated words. If the participant responded that they knew the answer upon the first presentation of a question, a list of seven unrelated words appeared so as not to alert the participant to the priming nature of the second task. To assess the effects of mode of production, half the participants read the words silently, and half read them aloud.

Figure 2. Flowchart of researcher’s key sequences during computer experiment.Figure 2. Flowchart of researcher’s key sequences during computer experiment.
Figure 2. Flowchart of researcher’s key sequences during computer experiment.


After the pronounceability task, participants were given another chance to answer the question that they did not know or were having a TOT for. This stage was critical to evaluate whether phonological primes aided TOT resolution. If they still could not answer the question, a partial information section allowed the participant to say any information they knew about the word, such as its first letter or number of syllables. After all 80 questions had been presented, participants completed the multiple choice recognition test.


Table 2
Percent of Target Resolution Following Primes and Unrelated Words After an Initial TOT Response
Initial Response Priming Condition
Mode of Production
Aloud
  
Silent
Mean
SD
Mean
SD
TOT  
(N=40)
  
(N=39)
First Syllable
35.5
33.7
50.2
33.9
First Phoneme
30.1
35.8
32.4
29.2
Unrelated
32.4
31.4
22.9
29.6


RESULTS

Table 2 and Figure 3 display the mean percentage of TOT resolution for each priming condition and mode of production. An analysis of variance (ANOVA) was conducted with production as the between- and condition as the within-participants factor. The results showed a main effect of condition, F (2, 154) = 6.09, MSE = .08, p < .003. Follow-up tests showed a greater percent of TOTs were resolved following first syllable primes (M = 42.8%) than either first phoneme primes (M = 31.3%), p< .016, or unrelated words (M = 27.7%), p< .002. There was no significant difference in resolution between first phoneme and unrelated conditions, p> .426. The ANOVA indicated no main effect of production, F < 1. There was a significant interaction between condition and production, F (2, 154) = 3.52, MSE = .08, p< .032. Follow up tests within each priming condition showed a marginally significant trend in the first syllable condition for greater TOT resolution following silent primes (50.2%) than aloud primes (35.5%), p< .057. There was no significant difference between aloud vs. silent for the first phoneme condition, F< 1 or the unrelated condition, F (1, 77) = 1.91, MSE = .09, p> .16. When examining differences in conditions within each mode of production, there were significant differences in silent primes, F (2, 154) = 9.14, MSE = .08, p < .001, but not in aloud primes, F< 1. Follow up tests showed that first syllables provided significantly more priming (50.2%) than first phonemes (32.4%), p < .01, or unrelated words (22.9%), p < .001. There was no significant difference between first phoneme and unrelated conditions for silent primes, F (1, 78) = 2.30, MSE = .15, p > .132.


Figure 3. Mean percent of resolution following TOT first response as a function of condition (first syllable, first phoneme, and unrelated) and production (aloud and silent).
Figure 3. Mean percent of resolution following TOT first response as a function of condition (first syllable, first phoneme, and unrelated) and production (aloud and silent).

Correlational analyses were then conducted to see if Nelson-Denny Vocabulary Score was related to the percentage of TOT resolution following each priming condition. There was a significant correlation between Nelson-Denny score and the percent of resolution following first-syllable primes, r(81) = .307, p < .005. In contrast, there was no significant correlation between Nelson-Denny score and TOT resolution following first-phoneme primes, r(82) = .136, p > .222, or following unrelated primes, r(82) = -.04, p > .723.

In terms of partial information retrieved during TOTs, participants reported a word that was phonologically related (i.e. sounded like) the target word 10% of the time. The most common specific type of phonological partial information that was related to the target answer was the first letter, provided for 8.4% of the unresolved TOTs. The second most common type of information was first syllable and number of syllables, each given on 2% of the items.

DISCUSSION

These data support the TDH, since TOT resolution following first syllable primes was greater than the unrelated condition, while providing evidence against the blocking hypothesis which predicted that phonological primes would hinder resolution (Jones & Langford, 1987, as cited in Schwartz, 2002). However, first phoneme primes did not facilitate TOT resolution compared to unrelated words. This suggests that first-phoneme and first-syllable phonology play fundamentally different roles in TOT resolution. Correlational analyses with vocabulary score also support this distinction, as only first-syllable priming increased as a function of better vocabulary. Perhaps participants who have strong vocabularies may have stronger connections between syllable nodes and lexical nodes, allowing them to reactivate first-syllable phonology more easily and retrieve the missing words.

Analysis of the partial information phase showed that the most common phonological information reported was the first letter. This suggests that first phonemes may be redundant and therefore ineffective primes that would not be expected to aid TOT resolution. Alternatively, perhaps a phoneme is simply not a large enough prime to aid resolution.

A significant interaction showed reading primes silently in the first syllable condition produced greater resolution than reading aloud. I noticed that some participants in the aloud condition seemed anxious to pronounce the words perfectly. Since some words were rather difficult to pronounce, those reading aloud may have become anxious about their performance, subsequently hindering TOT resolution. In any case, the extension of phonological priming effects to silent reading suggests that a change in connection strength at the phonological level, independent of nodes required for overt production of speech, is sufficient for facilitating word retrieval.

This research serves to add support to the Transmission Deficit Hypothesis (Burke et al., 1991), showing that first syllable priming can strengthen neural connections between semantic, lexical, and phonological nodes resulting in greater resolution of TOTs. The data also suggests how TOTs may be resolved in everyday life. For example, since reading primes silently aids TOT resolution, this may help to explain how the answers to TOTs seem to “just pop into our heads”.

REFERENCES

  1. Arana, Abrams, L., & White, K. K. (2001). Isolating phonological components that increase tip-of-the-tongue resolution. Manuscript in preparation.

  2. Burke, D. M., MacKay, D. G., Worthley, J. S., & Wade, E. (1991). On the tip of the
    tongue: What causes word finding failures in young and older adults? Journal of Memory and Language, 30, 542-579.

  3. James, L. E., & Burke, D. M. (2000). Phonological priming effects on word retrieval and  tip-of-the-tongue experiences in young and older adults. Journal of Experimental  Psychology: Learning, Memory, and Cognition, 26 (6), 1378-1391.

  4. Schwartz, B. L. (2002) Tip-of –the-tongue states: Phenomenology, mechanism, and lexical retrieval. Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

  5. White, K. K., & Abrams, L. (2002). Does priming specific syllables during tip-of-the-tongue states facilitate word retrieval in older adults? Psychology and Aging, 17 (2), 226-235.


Author's Note:
This research is part of a larger article that will be published in the psychology journal, Memory & Cognition.

--top--

Back to the Journal of Undergraduate Research