Predictors of Lexical Accessibility of Common and Proper Nouns in Older Age: Evidence from the Tip-Of-The-Tongue State

City University of New York (CUNY) CUNY Academic Works

Dissertations, Theses, and Capstone Projects CUNY Graduate Center

6-2021

Predictors of Lexical Accessibility of Common and Proper Nouns in Older Age: Evidence from the Tip-of-the-Tongue State

Amy Victoria Vogel-Eyny The Graduate Center, City University of New York

How does access to this work benefit ou?y Let us know!

More information about this work at: https://academicworks.cuny.edu/gc_etds/4367 Discover additional works at: https://academicworks.cuny.edu

This work is made publicly available by the City University of New York (CUNY). Contact: [email protected]

PREDICTORS OF LEXICAL ACCESSIBILITY OF COMMON AND PROPER NOUNS IN OLDER AGE: EVIDENCE FROM THE TIP-OF-THE-TONGUE STATE

AMY VOGEL-EYNY

A dissertation submitted to the Graduate Faculty in Speech-Language-Hearing Sciences in

partial fulfillment of the requirements for the degree of Doctor of Philosophy,

The City University of New York

2021

AMY VOGEL-EYNY

Predictors of lexical accessibility of common and proper nouns in older age:

Evidence from the tip-of-the-tongue state

Amy Vogel-Eyny

This manuscript has been read and accepted for the Graduate Faculty in Speech-Language- Hearing Sciences in satisfaction of the dissertation requirement for the degree of Doctor of Philosophy.

______Date Loraine K. Obler, Ph.D.

Chair of Examining Committee Date

______Date Mira Goral, Ph.D., CCC-SLP

Executive Officer

Supervisory Committee:

Elizabeth Galletta, Ph.D., CCC-SLP

Mira Goral, Ph.D., CCC-SLP

External Reviewer:

Jamie Reilly, Ph.D., CCC-SLP

THE CITY UNIVERSITY OF NEW YORK

iii

ABSTRACT

Predictors of lexical accessibility of common and proper nouns in older age:

Evidence from the tip-of-the-tongue state

Amy Vogel-Eyny

Advisor: Loraine K. Obler, PhD

One of the notable language difficulties experienced by healthy older adults is word retrieval failure, specifically the tip-of-the-tongue state (TOT). A TOT occurs when one has a strong sense of knowing the word, such that the semantic content is accessed, but the entirety of the word’s phonology is temporarily inaccessible. Such retrieval difficulty is attributable, at least in part, to characteristics of the target word. Psycholinguistic features may uniquely influence the semantic and/or phonological stages of word production. An additional factor known to influence TOT-likelihood is noun type: proper nouns elicit TOTs more often than do common nouns. The discrepancy between the likelihood of a TOT for the two noun types is hypothesized to be due to their differential representation in the mental lexicon. The difference hinges on the connection architecture at the semantic level (between semantic nodes and the lemma node) for common and proper nouns – the former characterized by convergent, many-to-one connections and the latter by one-to-one connections. The extent to which the representation of common and proper nouns accounts for retrieval difficulties as a consequence of psycholinguistic factors known to interact with the semantic level and phonological levels is poorly understood.

Therefore, this dissertation examined the contribution of several psycholinguistic features to the likelihood of successful retrieval at the semantic and phonological stages in a set of common and proper nouns.

Fifty-two monolingual English-speaking, healthy older adults between the ages of 54 and

89 participated in a TOT-inducing, computerized word naming task. Participants named targets from a selected subset of the total stimulus set of 1,102 words (587 proper nouns and 515 common nouns). Each target was cued independently from a picture and definition; however, cue type was counterbalanced across participants such that no-one saw the same target in both cue modalities. Analyses focused on the influence of the psycholinguistic features (namely self-rated frequency and familiarity, Zipf frequency, MRC familiarity, word-length in phonemes, neighborhood density, and first-syllable frequency) on word-retrieval performance at both stages of retrieval using a two-step model of TOTs.

The results of the current research offer novel evidence for the independent influence of frequency and familiarity on the likelihood of retrieval success at the semantic and phonological levels of lexical processing. Specifically, frequency was found to benefit both stages of retrieval for proper names and the phonological stage alone for common nouns. This finding suggests a frequency-related advantage such that one-to-one connections are favored at the semantic and phonological levels. By contrast, familiarity benefited retrieval at both stages for both noun types, indicating that an effect of familiarity is agnostic to connection architecture at each level.

The present study also offers new evidence for a cue-related retrieval advantage at the semantic level for common nouns (but not proper nouns) retrieved from picture cues. There was no effect of the other psycholinguistic features tested for either word type or stage of lexical processing.

This research establishes the independent contribution of psycholinguistic features to TOT occurrence in common and proper nouns in relation to the overall architecture of the semantic and phonological systems.

Steven Louis Vogel

There is a light that never goes out.

ACKNOWLEDGMENTS

This odyssey that is graduate school, punctuated by some of the highest-highs and lowest lows, illuminated, enriched, and shaped my nature – much to the better. There are many individuals who are owed my fondest appreciation, and my eternal thanks for their encouragement, support, and kindness.

To Loraine Obler, my advisor and trusty captain, who had taken me into her laboratory as a volunteer and had strengthened my resolve to pursue my doctorate in Speech-Language-

Hearing Sciences: thank you for helping me to realize the necessity and importance of reaching beyond one’s comfort to acquire new skills and new experiences, and to understand that to do so unapologetically is even better. There is thoughtfulness and care in all you say and do – from friendly conversations about daily happenings to complex discussions of methodological approaches – that I hope to carry with me. Your passion for language in aging became my own.

To Elizabeth Galletta, my committee member and faithful guide, who brought me into the folds of her research: to you I owe my conversion to a clinician researcher. With grace, you brought humanity into the laboratory and into the classroom – reminding those around you of the people behind the conditions we speech-language pathologists and researchers ultimately serve.

Thank you for ceaselessly advocating for me over the years and for always believing that I deserved that. I wish to always see others as you do and to remain cognizant of the beauty of our work as clinician researchers.

To Mira Goral, my committee member and grounding force, who was consistently there to remind me of the ultimate goal of having started this doctorate: finishing it – you welcomed me into your laboratory and allowed me to play a small but meaningful role, to me, in overseeing

vii

some of the brightest and most hardworking students I have ever encountered. I thank and value you for your calm and rational approach to research and life more generally.

To current and former Neurolinguistics Laboratory members, an epic would not be complete without a group of fellow wanderers to make the return journey alongside. Thank you all for filling these intervening days with downright hysterical moments that nicely complemented downright unpleasant moments when your friendship and intellect were sought.

Special thanks and appreciation are due to dear friends and lab mates, Sameer Ashaie, a kindred soul, Aviva Lerman a keen motivator, Jet Vonk, an enthusiastic mentor, Iris Strangmann, a tireless supporter, Marta Korytkowska, a steadfast companion, and Taryn Malcolm, a fearless fellow dissertator.

To my research assistants, Natalia Dos Santos Salles, Daniela Castillo, Johanna Pino

Grisales, and Mairenit Liriano, whose efforts on this project not only allowed the study to come to fruition but also brought new insight and perspective to the research process. Collectively you always kept me on my toes and never allowed me to take an idea for granted, thank you.

To the Eyny family, who never anticipated another daughter, sister, or aunt, but allowed me the privilege of holding such dear positions in their lives. To be supportive, nurturing, and generous with your love is reflexive and instinctive. I have been fortunate in that I never needed to ask for help because you had already given it to me. With the utmost gratitude: Miriam and

Zvi, thank you for modeling unconditional love; Sharon and Natalie, thank you for remaining by my side with humor and affection; and Daniel, thank you for being a sincere human – untainted by the world, which feels rare these days.

To my sisters, Sarah and Jane, who were there from the beginning. Thank you for exemplifying characteristics of strength, determination, and perseverance against all odds.

viii

To my husband, Yaniv Eyny, my Penelope in jest, thank you for staying the course with me. There are many words of gratitude for you and for what you represent in my life that would outstretch this dissertation. I look up to you beyond measure, with your effortless, boundless, and selfless ability to support and guide those around you. To the question “ Do I dare?” your answer to me has unfailingly been “yes.” For that, for gently raising me up, and for making everything worth it, I love and thank you. Then there is Mina Eyny, my other bright spot in this world.

Thank you for your natural enchantment and your pureness of heart, qualities that I celebrate in you and aspire to in myself. This world and my life are more beautiful and more meaningful for having you both in it.

TABLE OF CONTENTS

List of Tables ...... xii List of Figures ...... xiii 1. Introduction ...... 1 1.1. The Tip-of-the-Tongue Phenomenon ...... 2 1.1.1. Inducing TOTs ...... 5 1.1.2. Measurement of TOTs ...... 8 1.1.3. Proper Noun Retrieval ...... 11 1.2. Lexical Features Influencing TOT Incidence ...... 16 1.2.1. Semantically-based Features ...... 17 1.2.2. Phonologically-based features ...... 23 1.3. The Current Research ...... 30 1.4. Hypotheses ...... 31 1.4.1. Specific Aim 1 ...... 31 1.4.2. Specific Aim 2 ...... 32 2. Methods and Materials ...... 35 2.1. Participants ...... 35 2.2. Materials ...... 36 2.3. Procedure ...... 40 2.4. Statistical Analysis ...... 44 3. Results ...... 52 3.1. Semantic stage lexical retrieval ...... 52 3.1.1. Full dataset (common and proper nouns) ...... 52 3.1.2. Analysis of Common Nouns ...... 55 3.1.3. Analysis of Proper Nouns ...... 57 3.2. Phonological stage lexical retrieval ...... 59 3.2.1. Overall analysis (common and proper nouns) ...... 59 3.2.2. Analysis of Common Nouns ...... 63 3.2.3. Analysis of Proper Nouns ...... 65 3.3. Analysis of Demographic Factors ...... 69 4. Discussion ...... 71 4.1. Overview ...... 71

4.2. Influence of cue modality ...... 72 4.3. Influence of noun type ...... 75 4.4. Semantic-level psycholinguistic factors ...... 76 4.4.1. Frequency effects at the semantic level ...... 76 4.4.2. Frequency effects at the phonological level ...... 78 4.4.3. Familiarity effects at the semantic level ...... 79 4.4.4. Familiarity effects at the phonological level ...... 80 4.5. Phonological-level psycholinguistic factors ...... 81 4.5.1. Phonological psycholinguistic factors at the semantic level ...... 81 4.5.2. Phonological psycholinguistic factors at the phonological level ...... 82 4.6. Aging effects ...... 85 4.7. Clinical Implications ...... 87 4.8. Limitations and Future Research ...... 88 4.9. Conclusion ...... 92 Appendices ...... 94 References ...... 161

LIST OF TABLES

Table 1 Demographic Characteristics ...... 35 Table 2 Correlation matrix of independent variables ...... 48 Table 3 Collinearity values for independent variables ...... 49 Table 4 Power analysis ...... 50 Table 5 Response Types Across Trials ...... 51 Table 6 Regression results for cue modality, noun type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of semantic stage failure ...... 53 Table 7 Regression results for cue modality, noun type, noun type*cue modality, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of semantic stage failure ...... 54 Table 8 Regression results for cue modality, subjective frequency, subjective familiarity, length-in-phonemes, first-syllable frequency, MRC familiarity, Zipf frequency, and phonological neighborhood density as predictors of common noun semantic stage failure ...... 56 Table 9 Regression results for cue modality, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of proper noun semantic stage failure ...... 58 Table 10 Regression results for cue modality, noun type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of phonological stage failure ...... 60 Table 11 Regression results for cue modality, noun type, noun type*cue type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of phonological stage failure ...... 62 Table 12 Regression results for cue modality, subjective frequency, subjective familiarity, length-in-phonemes, first-syllable frequency, MRC familiarity, Zipf frequency, and phonological neighborhood density as predictors of common noun phonological stage failure ...... 63 Table 13 Regression results for cue modality, subjective frequency, subjective familiarity, length-in-phonemes and first-syllable frequency, as predictors of proper noun phonological stage failure ...... 65 Table 14 Summary of Direction of Independent Variables in Full Models ...... 67 Table 15 Summary of Direction of Independent Variables in Common Noun Models ...... 68 Table 16 Summary of Direction of Independent Variables in Proper Noun Models ...... 69

xii

LIST OF FIGURES

Figure 1 TDH Account of Common and Proper Noun Representation ...... 4 Figure 2 Response Types Relative to Stage of Lexical Processing ...... 11 Figure 3 Two-Step Account of Word Retrieval ...... 25 Figure 4 Frequency Rating Scale for Common Nouns ...... 42 Figure 5 Frequency Rating Scale for Proper Nouns ...... 43 Figure 6 Familiarity Rating Scale for Common Nouns ...... 43 Figure 7 Familiarity Rating Scale for Proper Nouns ...... 44 Figure 8 Determination of Success or Failure at Two Stages of Lexical Processing ...... 47

xiii

1. Introduction

Healthy aging is marked by decline in selected language functions while other linguistic and cognitive areas remain relatively spared and perhaps even improve due to older adults’ lifetime of language use (e.g., Kavé, Knafo, & Gilboa, 2010). Lexical retrieval is one such linguistic domain that declines with age, and it refers to the process by which a word is accessed, typically for production. The inability to find an intended word often occurs in the everyday speech of older individuals. Indeed, aging research suggests that the ability to recall words that were once familiar becomes progressively difficult within the course of healthy aging (Burke & Shafto,

2008; Evrard, 2002; Goral et al., 2007; Kavé et al., 2010). Further, retrieval difficulty is markedly greater for proper nouns than for common nouns (Burke et al., 1991; Fogler & James,

2007; Ossher, Flegal, & Lustig, 2013).

Several studies have shown that the naming difficulties that emerge along the natural course of aging are due to impairments to the retrieval mechanisms rather than the deterioration of word knowledge (i.e., semantic, phonological, orthographic). More specifically, the repository for storing knowledge of words (mental lexicon) increases well into adulthood, while the control mechanisms for retrieving words from the storage area gradually decline with age (Kavé et al.,

2010; Newman & German, 2005). Evidence for this is observed in studies that indicate that the semantic system appears to remain relatively intact and even develops with age, which is evidenced by vocabulary increases across the lifespan. By contrast, access to phonological representations declines with advancing age (Bowles & Poon, 1985; Burke & Shafto, 2004;

Gollan & Brown, 2006; Goral, Spiro, Albert, Obler, & Connor, 2007; Mortensen et al., 2006;

Nicholas, Obler, Albert, & Goodglass, 1985; Shafto, Stamatakis, Tam, & Tyler, 2009).

The field of language production in aging has been extensively examined for several decades and continues to be investigated with the intent of understanding the locus of lexical retrieval difficulties and the factors that modulate performance. With this purpose in mind, the present investigation is concerned with the lexical system of healthy older adults and the lexical features that act on the system to either facilitate or interfere with word retrieval.

1.1. The Tip-of-the-Tongue Phenomenon

Models of lexical retrieval, for all their differences, appear to agree on the idea that word naming occurs in two stages: the accessing of word meaning followed by word form (e.g., Burke &

Shafto, 2004; Gollan & Brown, 2006). These models contain networks of interconnected nodes representative of the semantic system and the phonological system. The semantic system holds meaning-based content as well as lexical-syntactic information (lemma), and the phonological system consists of word sounds and spellings (lexeme) (Burke & Shafto, 2008). During word naming, the semantic network is primed as neuronal excitation spreads along many connections from semantic representations to the intended lexical node, converging on the target lexical representation. This process is articulated in a language-processing, connectionist model known as Node Structure Theory (e.g., Burke et al., 1991) wherein linguistic units are represented as nodes. The Transmission Deficit Hypothesis (TDH), which builds from Node Structure Theory, offers specific predictions about language production in healthy aging that relate directly to the reduced transmission of priming between stages of language production (Burke, MacKay,

Worthley, & Wade, 1991; Mortensen et al., 2006). More specifically, proponents of the hypothesis put forth that the strength of the connections between the various stages of language production, for instance between word meaning and word form retrieval, show age-related weakening. Lexical retrieval performance therefore is dependent on the efficiency of priming

across these connections. During word naming, transmission of priming must extend along many connections from semantic representations to the intended lexical node-- all of which converge on the target lexical representation. Thus, if, for example a specific concept had a weakened connection to the intended lexical node (e.g., “used for calculating” in the example given above), other concepts such as “sliding beads” would still be able to transmit priming to the target lexical item. By contrast, the transmission of priming from a specific lexical node to the corresponding phonological nodes follows one-to-one connections, which are more vulnerable to transmission deficits (Burke et al., 1991; Burke & Shafto, 2004; Burke & Shafto,

2008; Mortensen et al., 2006). See Figure 1. Reduced transmission of priming from lexical-to- phonological nodes is best exemplified by the phenomenon known as the tip-of-the-tongue state.

The most telling evidence that advancing age is marked by impaired phonological retrieval, and not deficient semantic access, has been reported in studies examining tip-of-the- tongue states (TOTs). A TOT is indicated by a “feeling-of-knowing” or, in other words, when the semantic and grammatical content of a word can be retrieved but only partial phonological information can be accessed, such as the number of syllables in the target word or the initial phoneme (e.g., cat starts with /k/). If the first stage is successful, but the second fails such that the semantic content does not activate the phonological representations sufficiently, then a TOT occurs (Foygel & Dell, 2000). TOTs are distinct from other types of speech errors because they are often difficult to recover from, as evidenced by retrieval being delayed such that the speaker either cannot proceed until the TOT is resolved or the speaker abandons retrieval of the target until it may be seemingly spontaneously accessed at a later time point (Gollan & Brown, 2006).

Figure 1

TDH Account of Common and Proper Noun Representation

Note. Figure reprinted with permission from Abrams and Davis (2016).

According to the TDH, TOTs are accounted for by semantic-to-phonological connections that are susceptible to breakdown with advancing age, which disproportionately affects lexical retrieval performance in older adults as compared to younger adults. Research in diary studies and laboratory studies utilizing picture naming and naming-to-definition tasks has shown that older adults are more likely than younger adults to experience TOTs (Burke et al., 1991; Evrard,

2002; Gollan & Brown, 2006; Gollan & Silverberg, 2001; Juncos-Rabadán, Facal, Rodríguez, &

Pereiro, 2010; Salthouse & Mandell, 2013; Shafto, Burke, Stamatakis, Tam, & Tyler, 2007;

Shafto, James, Abrams, Tyler, & Cam-CAN, 2017). Indeed, healthy older adults self-report that

TOTs are their most frustrating everyday naming failure (Ossher, Flegal, & Lustig, 2013). A contributing explanation for why TOTs may be more common for older adults than younger adults is that older adults’ rich vocabularies allow for more opportunities for TOTs, and this is evidenced in the performance of older adults as they tend to have both more correct retrievals and TOTs on tests of lexical retrieval than younger individuals (see Gollan & Brown, 2006 for a review). Therefore, the study of TOTs is an important area of investigation because it is an ecologically-relevant language-related difficulty experienced by healthy elders.

1.1.1. Inducing TOTs

The majority of studies have chosen either naming-to-definition or picture naming tasks for inducing TOTs; however, few studies to date have compared whether using descriptions or pictures to cue target retrieval will have a disproportionate impact on the quantity of semantic and/or phonological retrieval failures during noun retrieval. Hanley (2011) has suggested that there is no reason to suspect a difference in retrieval output following the two cue types based on similar TOT rates for identification of proper nouns from faces (Hanley & Chapman, 2008) and definitions (Hanley, 2011). Nevertheless, it is unclear whether the target stimuli in both tasks were the same in both studies (i.e., “book” presented in both picture and definition form) -- a necessary methodological consideration when comparing the two modalities, and only proper nouns were evaluated, not common nouns.

In the common noun literature, Brown and Nix (1996) found that, among older adults, when the same nouns were presented either as pictures or definitions (the modality of presentation counterbalanced across participants) TOT rates for picture-elicited nouns (56.7%) and definition-elicited nouns (54.7%) were statistically equivalent. The study results, however, do not unequivocally show that the two cue modalities are equivalent, since the authors do not

specify the psycholinguistic properties of noun stimuli utilized (i.e., frequency, familiarity, neighborhood density, etc.). The researchers only point out that the words had to be imageable, since the targets were provided both as a definition and as a picture. It is possible that the psycholinguistic features of the stimulus set could influence the TOT-inducing capacity of the cueing modality. For example, the authors appeared to have used low frequency, difficult words, and, as such, the tasks are performed with similar TOT rates. Perhaps systematically varying the level of psycholinguistic characteristics (i.e., frequency, neighborhood density, familiarity, etc.) of the common noun stimuli would have had a different effect on retrieval performance for the two cue types. As well, there were only 50 stimuli, which calls the robustness of the effect into question having relatively few targets to base the finding on.

In a study conducted by Read and Bruce (1982), the authors examined proper noun retrieval from picture and definition cues among a group of presumably young adults (the study makes no reference to the age of the participants, but note that the population came from several universities so we may assume the majority of participants, if not all, are college aged). Similar to Brown and Nix (1996), half of the 204 items were cued from pictures and the other half from definitions, and this was counterbalanced such that all items were tested in the two modalities.

The researchers found that summed TOT rates across participants for both definitions (201

TOTs) and pictures (199 TOTs) were almost identical. The protocol, however, was flawed because it required at least one TOT every session on each of the definition- and picture-cued items, and if one modality did not elicit a TOT, items within that modality would be presented until a TOT state was achieved. This uncommon procedure of adding trials within a session until a TOT state occurred, may have artificially inflated the TOT rate. As well, the authors calculated TOTs using the raw occurrence summed across participants, whereas Brown and Nix

(1996) calculated TOTs for each cue as a percentage of noncorrect retrievals (i.e., TOTs/TOTs +

Don’t Knows). From the few studies that have been conducted, differences in methodologies for eliciting TOTs and determining TOT incidence make it difficult for strong conclusions to be drawn about cueing efficacy and how often TOTs occur.

Research involving disordered populations, such as people with epilepsy and aphasia, have found modality-specific impairments, wherein patients showed poorer performance in either naming-to-definition or picture-naming tasks. A 2008 study conducted by Zamarian and colleagues examined the retrieval of proper nouns from picture and definition stimuli in a group of individuals with left or right temporal lobe epilepsy and found that the groups differed in retrieval performance on picture naming (individuals with left temporal lobe epilepsy performing more poorly than those with right) but not on naming-to-definition. Turning to individuals with aphasia, response latencies were found to be longer in a naming-to-definition task relative to picture naming (Goodglass & Stuss, 1979). Similarly, these findings suggest that presenting cueing from different stimulus modalities may involve distinct retrieval processes. Thus, there is evidence to suggest that cue modality influences lexical retrieval. Taken together, an evaluation of whether TOT occurrence is dependent on the nature of the task is warranted given the conflicting findings in the literature. Specifically, the heterogeneous methods utilized in prior

TOT studies limits what can be known about the effect of picture and definition cues on TOT incidence.

It is of interest to understand whether naming-to-definition as compared to picture naming results in differences in semantic and phonological access to proper and common nouns in aging. Specifically, researchers employ both task types interchangeably when assessing retrieval of nouns and compare results across studies with the assumption that TOT rates are

similar among stimulus modalities for both word classes, which may not be the case. The two studies that have been conducted on definition and picture cue comparisons for common (Brown

& Nix, 1996) and proper nouns (Read & Bruce, 1982) have flawed methodologies relating to the number of stimuli, the paradigm for eliciting TOTs, and the calculation of TOT incidence. A better understanding of the TOT rates for common and proper nouns given definition and picture cueing modalities will assist future researchers in selecting a methodology for eliciting TOTs and in determining the extent to which older adults experience retrieval difficulties for common and proper nouns.

1.1.2. Measurement of TOTs

There are a few methods that researchers have employed to compute TOT incidence relative to overall retrieval performance. One approach to the measurement of TOT data is to examine

TOTs from the perspective of unsuccessful retrievals – the proportion of TOTs out of all non- successful retrievals (e.g., Burke et al., 1991):

TOTs

(N − GOTs)

In this equation a successful retrieval is termed a GOT, as in the participant retrieved the target correctly. The rationale behind this error-centered perspective on TOTs is that a trial that results in a successful retrieval could not have produced a TOT (i.e., GOTs obviate the potential for

TOTs). However, according to Gollan and Brown (2006), studies utilizing this type of TOT calculation tend to show more robust age effects on TOTs because older adults often have a higher rate of correct retrievals than younger adults as a result of their larger vocabularies, leaving fewer opportunities for a TOT to occur. Additionally, Gollan and Brown (2006) have proposed that raw TOTs do not provide a sufficient account of group differences because an

increase in raw numbers of TOTs leave the impression that older adults are disadvantaged compared to younger adults.

In order to correct for these limitations in TOT measurement, Gollan and Brown (2006), compute TOTs within the framework of current language production models. They observed that all such models share two essential features: 1. semantic access occurs first (referred to as Step

1), and 2. phonological access occurs second (referred to as Step 2). As well, characteristics of

TOTs in aging, such as access to partial phonological information and the occurrence of competing alternatives, provides additional evidence for the locus of the impairment arising during phonological retrieval (Burke & Shafto, 2004). As previously mentioned, TOTs may occur despite the availability of partial phonological information. Gollan and Brown (2006) suggest that partial phonological retrieval is actually indicative of better functioning retrieval processes compared to no retrieval at all. That is, the authors argue that according to current language production models, greater TOTs in older adults may indicate partially successful retrieval, since they reflect complete access to semantic representations (referred to as Step One) and incomplete retrieval of the word form (referred to as Step Two). A TOT, from this perspective, is the result of better retrieval processes than a completely failed retrieval.

With this in mind, the authors propose a set of calculations that take into account the types of responses an individual can make during word retrieval and how these responses are reflected in the two-step model of TOTs. The authors determined Step One failures by dividing all responses that were neither positive TOTs (i.e., TOTs for the target item; which the authors call ‘+TOTs’) nor correct retrievals (i.e., successful retrieval of target; noted by the authors as

GOT) by total number of “N” targets. Both +TOT’s and GOTs indicate successful Step One retrieval (which implies access to semantics), hence their exclusion from other production types

in the semantic failures analysis. However, as noted by Juncos-Rabadán et al. (2010), this is not a valid measure of semantic access failure due to the inclusion of words that were not known to the participant in both the numerator and denominator. As a modification, I propose a calculation that additionally excludes words that were never known, which I refer to as Never Knews (NK), as retrieval of these targets would not have been possible at either stage of production:

n − (+TOTs + GOTs + NKs)

Step Two failures are indicated by the division of positive TOTs by the number of positive TOTs and GOTs:

+TOTs

+TOTs + GOTs

Positive TOTs are the only example of a production that includes successful retrieval of semantics but failed retrieval of complete phonology. These calculations consider TOT occurrences alongside other possible responses in a naming task, which allows for a complete characterization of lexical retrieval performance. This fine-grained production analysis will allow for a complete characterization of retrieval performance by taking into account whether speech production is the result of a Step One failure (failed semantic access) or Step Two failure

(failed phonological access). See Figure 2. This procedure is outlined in detail in the methods section below. Gollan and Brown (2006) showed that older adults were more likely to experience successful Step One retrieval compared to younger adults, suggesting an age-related retrieval advantage for semantic access. However, no such age-related benefit was observed for step two retrieval.

Figure 2

Response Types Relative to Stage of Lexical Processing

Note. Response types are shown relative to success or failure in completing 2 retrieval steps: semantic and phonological. Not GOT indicates semantic stage retrieval failure. +TOTs imply phonological stage retrieval failure. Correct retrieval (GOT) suggests successful access at both stages.

1.1.3. Proper Noun Retrieval

A substantial body of research focuses on proper name retrieval as compared to retrieval of other word types, largely due to the disproportionate difficulty older adults experience retrieving proper nouns (Burke et al., 1991; Fogler & James, 2007; Ossher, Flegal, & Lustig, 2013).

Empirical research validates the experience of older adults, suggesting that proper names become

more difficult to retrieve over the lifespan and more often result in retrieval deficits relative to common nouns, as evidenced by increased TOTs, longer response times, and lower accuracy

(Burke et al., 1991; Evrard, 2002; James, 2006; Fogler & James, 2007; Juncos-Rabadán et al.,

2010; McWeeny, Young, Hay, & Ellis, 1987; Oberle & James, 2013; Reason & Lucas, 1984;

Salthouse & Mandell, 2013). Furthermore, younger adults are not exempt from this experience as they too show a higher rate of retrieval failure for proper nouns than other word types, though the magnitude of the effect may differ compared to older individuals (e.g., Burke et al., 1991;

Evrard, 2002). Although the majority of studies have found that older individuals experience more TOTs than young adults for proper names, Oberle and James (2013) did not observe this effect but did show that younger adults more accurately retrieved proper noun targets than older adults. There is no clear explanation for the discrepancy in results, though methodological differences are rampant among studies with respect to both stimulus selection and experimental procedures.

Research into the variables influencing proper noun retrieval that may contribute to the observed word naming difficulty compared to other word types has been conducted. Prior literature has suggested that there are a number of variables that can influence proper name retrieval (McWeeny et al., 1987; Fogler & James, 2007). More specifically, proper names tend to be less familiar, of lower frequency, less meaningful, and less imageable. With respect to reduced imageability, proper nouns typically do not offer information on the semantic attributes of the referent (e.g., physical features) (Abrams & Davis, 2016; Brédart, 2017; Cohen, 1990;

Fogler et al., 2010). To echo Cohen & Burke (1993), people carry semantic information, people’s names do not. As well, proper nouns contain more words (e.g., Catherine Zeta Jones vs. silo) and a greater variety of possible phonological sequences (i.e., novel syllables or novel

combinations of familiar syllables) than common nouns (Brennen, 1993). As an example of the latter, Brennen (1993) suggests that one’s reaction to the word ‘dreaner’ depends on its specification as a common or proper noun. Indeed, one would not be surprised at all to hear of a surname characterized by unfamiliar phonology (such as ‘dreaner’); however, one would be surprised if told that the person’s occupation was a ‘dreaner,’ and one might think they misheard the word. Interestingly, this range of plausible phonology for proper nouns has increased over the last century (Ramscar et al. 2014).

Several researchers have addressed the meaninglessness or nondescriptive nature of most proper names, which has been proposed as one reason why proper nouns are so difficult to retrieve. Proper nouns are often considered semantically unique or nondescriptive since they do not contain representative information about the identity or mental and physical attributes of the referent (Semenza 2011; Waldron, Manzel, & Tranel, 2014). That is, research suggests that the difference is that proper nouns refer to individual entities reliant on unique semantic properties, and do not rely on a set of attributes to the same extent as common nouns do. Proper nouns appear to refer to unique entities (i.e., attributes are specific to a given person or place) and common nouns to categories (Gorno-Tempini & Price, 2001; Ross & Olson, 2011). Described another way, proper nouns have “token reference” (i.e., instantiation/instance), and common nouns have “type reference” (i.e., classification/category) (Semenza, 2006; 2011). Thus, according to Semenza (2006), the relationship between a proper noun lexeme and its lemma (its reference) is weakened by the lack of an interrelated web of semantic attributes, which may account for the disproportionate difficulty in proper noun retrieval.

A number of studies to date have attempted to control for factors that may contribute to the disproportionate difficulty in retrieving proper nouns relative to common nouns (Festini,

Hartley, Tauber, & Rhodes, 2013; Hanley, 2011; Fogler & James, 2007; McWeeney et al.,

1987). Brédart (1993) suggested that proper nouns are more difficult to retrieve than common nouns because they lack an acceptable alternative label. Hanley (2011) found a difference in common noun and proper noun retrieval when the stimulus set was matched for familiarity; however, they also observed more incorrect responses for common nouns than proper nouns, consistent with the former’s likelihood of eliciting synonyms. To test Brédart’s claim, the author designed a second experiment wherein proper and common nouns were matched for familiarity as well as “don’t know” responses and number of alternative correct responses. In support of

Brédart’s view, the researchers found no difference in TOT occurrence in common and proper nouns. Thus, familiarity alone does not appear to account for the difference in retrieval difficulty for common and proper nouns. However, the researchers employed two different methodologies to elicit TOTs; experiment one involved naming from picture cues and experiment two from definition cues. It is unclear whether cue type influences TOT occurrence; therefore, it is challenging to interpret the author’s finding, since the two studies are not comparable. Additional research is needed in order to find support for the conclusion drawn by Hanley (2011) with regard to the role of circumlocution in retrieval of both word types.

In a study conducted by Fogler and James (2007), participants named famous cartoon characters that had a descriptive (e.g., Snow White) or nondescriptive (e.g., Charlie Brown) name. Consistent with prior research, older adults performed less accurately in proper name retrieval than younger adults for both descriptive and nondescriptive names. Among the older individuals, participants performed significantly better when retrieving descriptive proper names relative to nondescriptive ones, which reduced the difference between younger and older adults in retrieval accuracy for descriptive proper names. These findings suggest that although older

adults were significantly worse at retrieval for descriptive and nondescriptive words compared to younger adults, descriptive names are significantly easier to retrieve than nondescriptive names in advancing age. Similarly, McWeeny et al. (1987) and Festini and colleagues (2013) attempted to equate proper and common nouns for factors that may lead to disproportionate difficulty in retrieval of the former, such as meaningfulness and frequency of use, in a group of young adults.

Consistent with other studies, McWeeny et al. (1987) found that proper names such as Mr.

Porter were retrieved less successfully by younger adults than same-word common nouns such as porter. Interestingly, Festini et al. (2013) observed that when proper nouns and common nouns are both nonwords, thereby controlling for the number of possible semantic associations, they are retrieved at the same rate.

Given that the semantic system for proper nouns (i.e., person-specific knowledge) is well-preserved with advancing age (Pistono et al., 2019), TOT occurrence is unlikely to result from an impairment at the semantic level. One hypothesis as to why proper nouns are disproportionately more difficult to name than other word types has to do with the architecture of lexical representations (Burke et al., 1991), which is consistent with the TDH. To exemplify the difference in retrieval of common and proper nouns, consider the semantic representation of the occupation noun “baker” and the representation of the family proper name “Baker.” The occupation noun receives summation priming from convergent semantic to lexical nodes (e.g., bakes bread, gets up early). Contrastively, the family proper name “Baker” does not receive convergent priming; rather semantic connections converge on the proper noun phrase “John

Baker,” and only a single divergent connection from the proper noun phrase to “Baker” exists, leaving it vulnerable to transmission deficit, which increases with age. See Figure 1. The TDH may also be the best account of the meaningfulness effect observed in prior research (Festini et

al., 2013; Fogler & James, 2007; McWeeny et al., 1987). Indeed, descriptive names might be easier to produce than nondescriptive ones because they have additional connections between the name and semantic nodes, such that descriptive names resemble nonproper name stimuli.

In sum, both younger and older adults alike experience lexical retrieval difficulties -- albeit older adults to a greater extent than their younger counterparts -- for proper nouns compared to other word types. Several factors may influence proper name production such as descriptiveness and frequency. Overall, the inordinate difficulty that people experience with proper names is best explained by the TDH within an interactive activation model that suggests that the representational architecture of proper names, which is distinct from that of common nouns, affects lexical retrieval since one-to-one connections are more susceptible to transmission deficit than are many-to-one connections.

Taken together, the literature suggests that lexical retrieval abilities decline with advancing age, and TOT states are the most common evidence of word naming failure – the occurrence of which is often linked to reduced access to proper nouns.

1.2. Lexical Features Influencing TOT Incidence

Research suggests that there is considerable variability in the propensity of lexical targets to induce TOTs (Abrams & Davis, 2016; Brown, 2012). To the extent that some lexical items are more or less susceptible than other items, one can ask whether a property or set of properties associated with such words can be determined in order to further elucidate the nature of lexical retrieval processing (Hodgson & Ellis, 1998; Newman & German, 2005). An understanding of the feature or set of features that influence TOT occurrence is often the express purpose of TOT researchers’ examination of incidence. Improved knowledge of such factors would better inform

TOT researchers in their selection and manipulation of target TOT words.

Models of lexical retrieval distinguish between semantic and phonological representations, and if such representations are structurally independent of one another then lexical factors can be conceived of in relation to such representations, particularly in terms of whether a specific factor is likely to impact one component of processing over another or even the connections between the two representations. This method of examining the impact of lexical factors on retrieval was proposed by Newman and German (2005); the authors posited that certain lexical factors (e.g., word frequency) are linked to either the semantic stage, semantic-to- form stage, or the phonological stage of lexical processing. In the upcoming sections, the influence of lexical factors on the stages of lexical retrieval will be discussed.

1.2.1. Semantically-based Features

Several lexical properties such as familiarity and frequency have been implicated in lexical retrieval research, and their interaction with one another has also been considered. In their 2005 paper, Newman and German suggested that both frequency and familiarity likely influence the transmission of information between semantic and form-based representations, since, the authors argue, these properties are related to the word’s use. This is in contrast to phonologically-based features, such as word length, which are not defined by use.

1.2.1.1. Frequency

Word frequency refers to how often a word appears in spoken or written language. TOT researchers have long-employed difficult (i.e., low frequency) common nouns in order to induce

TOTs, since the prevailing belief has been, and perhaps continues to be, that TOT occurrence is uncommon for high-frequency word retrieval (e.g., Brown & McNeill, 1966; Burke et al., 1991;

Gollan & Brown, 2006; James & Burke, 2000). Over the years, diary studies of naturally- occurring TOTs have generally drawn this conclusion (Burke et al., 1991; Ecke 2004). Burke

and colleagues (1991) examined naturally-occurring TOTs among younger and older adults and found that approximately half of participants’ resolved TOT words were so infrequent as to be unlisted in referenced norms (i.e., Francis and Kucera, 1982). Furthermore, the median TOT frequency of the resolved common nouns (abstract and object names) that existed in Francis and

Kucera’s corpus was lower than the median frequency of the open class words in the corpus itself (84 occurrences per million), with only 11% of the words Burke and colleagues (1991) listed falling above that median. Perhaps most interestingly, however, is the marginal trend wherein older adults experienced more of these uncatalogued TOTs than did younger adults.

These results indicate that the frequency effect in TOT occurrence may be modulated by age, with low frequency words causing more TOTs in older than younger adults (Abrams & Davis,

2016). One explanation as to why older adults demonstrate more TOTs for low frequency items may be that they have more opportunities for a TOT to occur given their greater knowledge and use of infrequent words. Indeed, Kavé et al., 2010 examined the relation between lexical frequency of noun usage in connected speech and aging and found an unexpected negative correlation, suggesting that with advancing age there is a greater quantity of low frequency common nouns retrieved relative to younger adults. Thus, older adults have a larger vocabulary than younger adults and have more low-frequency words in their mental lexicon from which to retrieve.

Of the few studies that have systematically compared TOT rates for low- and high- frequency common noun targets among younger and older adults, the findings support the frequency effect observed in the diary studies mentioned above. In two experiments with college-aged young adults, Harley and Bown (1998) found significantly more TOTs for low- frequency than high-frequency targets retrieved from definition cues, though the percent

difference was relatively small (experiment one: 3%; experiment two: 1%). This pattern of performance has also been observed among older adults (Vitevitch & Sommers, 2003). More surprising, however, is the finding of Astell and Harley (1996) that in their group of older adults low frequency words alone accounted for all TOTs, though the stimuli consisted of a mere 24 words: 12 high frequency and 12 low frequency. Researchers taking a correlational approach to the question of frequency have also observed a negative association between word frequency and

TOTs (Gollan & Silverberg, 2001; Gonzalez, 1996).

Few studies to date have examined the effect of frequency on TOT rates during proper noun retrieval. Brédart (2017) explains that this might be due to the fact that objective frequency counts for proper nouns simply do not exist in American English as they do for common nouns, though such measures are found in other languages: French (Bonin, Perret, Meot, Ferrand, &

Mermillod, 2008) and British English (Smith-Spark, Moore, Valentine, & Sherman, 2006). In

Bonin and colleagues (2008) study, college-aged students rated how frequently they heard, read, or produced the names of French celebrities, and consistent with objective ratings, there was a negative correlation between frequency and TOT rate.

Taken together, low-frequency words are disproportionately more susceptible to TOT states than high-frequency targets. The TDH predicts that reduced semantic-to-phonological transmission of priming for infrequent words is the root cause. More specifically, a word’s frequency is directly related to its likelihood of sending activation from lemma representations to corresponding phonological representations. In this way, the semantic-to-phonological connections for infrequently occurring lexical targets are weaker since they are not often utilized.

According to the two-step method for calculating TOTs, a frequency effect would be evident in an increase in Step 2 failures.

Although the frequency effect has been observed in research on common and proper nouns, to my knowledge, no study has examined the effect of word frequency on retrieval of common and proper nouns within the same experimental paradigm. Proper nouns are generally lower in frequency than are common nouns; however, common nouns can be infrequent as well.

Brédart (2017) point outs that low frequency may be a factor common to proper nouns but not necessarily unique to them.

1.2.1.2. Familiarity

Researchers have argued for the importance of examining other lexical factors, both independent from and in addition to frequency, that might modulate and interact with TOT occurrence in aging (e.g., Farrell & Abrams, 2011; Gollan & Brown, 2006; Hanley, 2011). Item familiarity, or how well known the item is to the speaker, is a psycholinguistic variable that may potentially influence lexical retrieval abilities in older age. A number of studies have found an effect of familiarity on lexical retrieval performance with age, such that items with greater familiarity are better retrieved than those targets with lower familiarity ratings (Burke et al.,

1991; Newman & German, 2005; Schmitter-Edgecombe, 2000). Diary studies examining familiarity of TOT target words have found that TOTs are more likely to occur when the target is highly familiar, and this holds true for both common and proper nouns (Burke et al., 1991;

Cohen & Faulkner, 1986; Ecke, 2004). It is worth noting that this finding is in contrast to word frequency, where high frequency words are less likely than low frequency words to result in a

TOT. As well, this effect of familiarity is particularly evident among middle-aged and older adults as compared to younger adults (Burke et al., 1991).

Burke and colleagues (1991) were interested in examining proper-name retrieval difficulties in order to understand the extent to which familiarity with proper names is impacted

with age. The authors examined whether familiarity with a name affects retrieval abilities. To investigate an effect of familiarity, the authors treated recency of acquisition of the proper name as a proxy measure for familiarity (i.e., proper names known for a long time tend to be more familiar than proper names that are newly learned). This approach is motivated by prior research, which suggests that older age is associated with greater difficulty recalling proper names

(McWeeny et al., 1987). In their diary study, Burke and colleagues (1991) found that proper names were more likely than any other word type to result in a TOT, and impaired access to proper names increased with age. Notably, newly-learned, less familiar proper names were not accounting for the effect; rather regardless of age, all participants experienced TOTs for proper nouns that they rated as being highly familiar and known for at least one year. In fact, for proper names that they experienced a TOT for, the average duration of acquaintance for older adults was 17.67 years. This finding implies that highly familiar proper names are more susceptible to

TOTs than are recently acquired proper nouns that may be less familiar. The authors propose that proper names are difficult to retrieve even when known for an extended time. Given that high familiarity is a pre-requisite for an opportunity to have a TOT, it would seem that familiarity alone does not contribute to the difficulty in retrieving proper nouns.

In contrast to findings from diary studies, laboratory investigations have found no such effect of familiarity on TOT occurrence. Hanley and Chapman (2008) manipulated proper noun familiarity (i.e., celebrities) and found no effect on the likelihood of TOT occurrence in their group of young adults (age range of 17-19 years). However, the trend was such that more TOTs were observed for low- than high-familiarity proper nouns. A possible explanation for why an effect of familiarity was observed in diary studies but not in Hanley and Chapman’s (2008) laboratory investigation may be that diary studies reflect day-to-day language use, which likely

includes the use of vocabulary already well-known to participants. In contrast, laboratory investigations include lexical items pre-defined by experimenters that do not account for a participant’s level of familiarity in advance of testing. Another possibility is that the researchers used too constrained a range of familiarity to allow for differences in TOT likelihood.

Employing a 1-5 scale where 1 represented “unfamiliar” and 5 “maximum familiarity,” the average rating for high familiarity celebrity names was 4.65 (range = 4.4-4.9) and low familiarity

3.84 (range = 3.4-4.3). I posit that these findings suggest a range, between both high and low familiarity, in which TOTs are more or less likely to occur.

The TDH provides support for a decline in transmission of priming between the semantic and phonological levels, such that the connections for words that are highly familiar are less likely to be vulnerable to age-related weakening. From a two-stage interpretation of word production, familiarity as a psycholinguistic construct affects both the semantic system as well as the phonological system. Knowledge of a lexical item will impact semantic access and may have negative downstream effects on retrieval of phonological content. Indeed, Hanley and Chapman

(2008) observed that both high- and low- familiarity of proper noun targets were associated with retrieval failures at both Step 1 and Step 2; however, the probability of a retrieval failure at both stages was more pronounced for low-familiarity relative to high familiarity words. Hanley

(2011) argued against the TDH in their study as there was no difference in common noun and proper noun retrieval when matched for familiarity and likelihood of eliciting alternatives.

However, their participants were young adults whose semantic-to-phonological connections are not subject to age-related weakening. As well, their targets were of relatively high familiarity, suggesting that the semantic content for the items may have equally benefitted both proper and common nouns such that retrieval deficits were not observed at the phonological level. Given the

discrepant findings between diary studies and laboratory investigations, additional research is crucially needed to further investigate the role of familiarity among healthy older adults during common noun and proper noun retrieval.

Frequency and familiarity are often treated as equivalent in TOT research; however, they likely exert a different effect on common and proper noun retrieval in aging. No study to date has manipulated both the frequency and familiarity of common and proper nouns within the same experimental paradigm. Although frequency and familiarity are positively correlated with one another, there are instances in which a word may be relatively low frequency and highly familiar

(Newman & German, 2005). Indeed, an additional consideration when examining the influence of lexical factors on TOT performance is that although the TDH assumes that proper nouns result in TOTs more often than common nouns, it does not exclude the possibility that factors interact

(e.g., frequency and familiarity) in such a way that might modulate the disproportionate difficulty in proper noun retrieval and perhaps even reverse it (Brédart, Brennen, Delchambre,

McNeill, & Burton, 2005).

1.2.2. Phonologically-based features

In addition to the existing body of research on the semantically-based features impacting lexical retrieval in aging, several studies have also been conducted on the influence of the phonological features of words on lexical retrieval processes across the lifespan (Gordon & Kurczek, 2013;

Hodgson & Ellis, 1998; Mortensen et al., 2006; Newman & German, 2005 ). The lexical properties proposed to affect the phonological system during word naming include word length, neighborhood density, and first-syllable frequency, which are the factors that have been examined in the literature to a greater or lesser extent. In the following sections, each lexical

factor will be considered in order to gain a deeper understanding of the properties of words that affect lexical retrieval.

1.2.2.1. Neighborhood Density

An oft-studied phonological property that appears to influence lexical retrieval, particularly in aging, is neighborhood density. It is defined as the number of lexical items that differ from the target production by an added, deleted, or substituted single phoneme (Gordon & Kurczek, 2013;

Newman & German, 2005; Vitevitch, 2003). For example, ripe has several phonological neighbors: pipe, ride, and rope. The TDH predicts, counterintuitively, that words with denser neighborhoods (i.e., more phonological neighbors) are less vulnerable to TOT states than are those from sparse neighborhoods (i.e., few phonological neighbors). This effect is explained by a spreading activation account that allows for feedback activation such that lemmas receive activation from the target’s phonological neighborhood (Abrams & Davis, 2016; Vitevitch &

Sommers, 2003). During retrieval of the word cat, for instance, the semantic concepts activate the lemma and transmit priming to its phonological nodes. In turn, the phonological nodes transmit feedback activation to lemmas that phonologically overlap with the target (e.g., rat, mat), and feedforward activation from those lemmas spreads back to the shared phonological nodes. See Figure 3 for a visualization of feedforward and feedback activation during the two- step process of lexical retrieval. In this way, the TDH predicts that feedback activation benefits words from denser phonological neighbors due to the strengthening of lemma-to-phonology connections, thereby reducing TOT likelihood.

The few empirical investigations into the role of neighborhood density have supported the predictions made by the TDH (Harley & Bown, 1998; Vitevitch & Sommers, 2003). In their group of young adults, Harley and Bown (1998) manipulated both frequency and neighborhood

density and found that words from sparse neighborhoods induced more TOTs than words from dense neighborhoods. The authors additionally observed an interaction that indicated the effect of neighborhood density was greatest for low frequency words. As such, the likelihood of TOT occurrence is greater for targets with sparse phonological neighbors and low frequency of use. In a second experiment, this finding held true when the authors controlled for the potentially confounding effect of word length. Specifically, sparse neighborhoods typically contain longer words. Taken together these findings have implications for increasing TOT states in that the co- occurrence of more than one target word dimension has an additive effect on TOT rate (Abrams

& Davis, 2016).

Figure 3

Two-Step Account of Word Retrieval

Note. Figure reprinted from Dell, Martin, & Schwartz (2007) with permission from Elsevier.

In their 2003 study, Vitevitch and Sommers investigated the influence of neighborhood density and neighborhood frequency on TOT rates for monosyllabic words. Neighborhood frequency is the average frequency of a target word’s phonological neighbors. The researchers

replicated the findings of Harley and Bown (1998): younger adults experienced more TOTs for low frequency and low neighborhood density targets. As well, there was no effect of neighborhood frequency among young adults. By contrast, there was no significant main effects of word frequency, neighborhood density or neighborhood frequency among older adults (mean age = 70.3); however, an interaction was observed wherein older adults exhibited more TOTs for words from low neighborhood frequency but only when the target was also low frequency and from a sparse neighborhood. Abrams and Davis (2016) explain the observed age effect for neighborhood frequency within a TDH framework, suggesting that age-related weakening of semantic-to-phonological connections reduce the feedback activation from low-frequency neighborhoods moreso than targets from high-frequency neighborhoods.

Despite these findings, additional research on the effect of neighborhood density on the

TOT rate of older adults is necessary. In Vitevitch and Sommers’ (2003) study the authors controlled for word and syllable length by utilizing monosyllabic words, which on average induced TOTs 2% and 3% of the time for younger and older adults, respectively. Given the very low TOT rates and the overall lack of word length variability in their stimulus set, further research on a broader set of common nouns is necessary to examine the effect of neighborhood density on TOT incidence in healthy elders.

1.2.2.2. Word Length

Although several studies have investigated the role of word length in naming performance, it is difficult to draw conclusions from those studies because the researchers did not examine whether word length independently influenced lexical retrieval (Mortensen et al.,

2006). That is, shorter words also tend to be of higher frequency, as put forth by Zipf’s Law, which states that there is an inverse relation between the number of occurrences of a word and its

length (Zipf, 1935). Indeed, prior studies that have found an effect of word length on lexical retrieval abilities may have reflected a sensitivity to a number of other variables as well.

Hodgson and Ellis (1998) matched for such factors known to correlate with word length, and they nevertheless found that there was a significant effect of word length (in phonemes) on naming accuracy among older adults, with shorter words having greater accuracy than longer words. This effect was only observed when naming occurred within 5 seconds of stimulus presentation. The researchers concluded that the rapid retrieval of targets is impeded by word length, and that this may be due to reduced activation from semantic-to-phonological representations, in particular for longer words that require greater input from the semantic system as they are more phonologically complex than shorter words.

Target-word length for proper names was shown to influence TOT rate among young adults (Hanley & Chapman, 2008). Specifically, the researchers examined two- and three-word celebrity names (e.g., Sean Penn and Billy Bob Thornton) and observed more TOTs for the latter than the former regardless of whether the targets were rated high or low familiarity. Utilizing the calculations suggested by Gollan and Brown’s two-step approach to TOTs (2006), the authors note that proper names composed of three names were more likely to result in a phonological- level deficit (at Step Two) than names consisting of only two words.

Short and long target-word length has also been measured by way of phoneme counts with similar results (Hanley & Vandenberg, 2010). Specifically, there was an effect of phoneme length such that long targets (composed of 6 or more phonemes) were more likely to induce

TOTs than short targets (composed of 4 or fewer phonemes), and this finding interacted with frequency such that long target words of low frequency elicited the greatest number of TOTs.

These findings can be explained within the TDH framework as follows: both more words and more phonemes are associated with an increase in phonological nodes that need activation, thus allowing for more opportunities for reduced transmission to occur and subsequent retrieval failures. Research on the effect of word length on TOT rate only included young adults (Hanley

& Chapman, 2008) or children (Hanley & Vandenberg, 2010). Older adults may be more susceptible to transmission deficits due to lemma-to-lexeme weakening than a younger demographic. Thus, there is a need to understand the influence of target length on TOTs in an aging population, as it is currently unestablished in the literature.

1.2.2.3. First-Syllable Frequency

Another dimension of TOTs that has been investigated for its influence on TOT incidence is first-syllable frequency, which is how often the initial syllable of a word is used in a language. According to Abrams and Davis (2016), high-frequency first syllables such as /di/ (as in dean, decoy) have an increased rate of occurrence in words. Contrastively, low-frequency first syllables include /ɒm/ (as in omelet and ombudsman), which only occasionally appear at the start of English words (Abrams & Davis, 2016). In a study of common noun retrieval, low-frequency first syllables alone contributed to TOT occurrence in both old-old (mean age = 80) and young- old (mean age = 68) adults compared to a group of college-aged young adults (Farrell & Abrams,

2011). Indeed, age differences in TOT rates disappeared during retrieval of high-frequency first syllables. These findings have been replicated and extended by Farrell (2012), such that low- frequency syllables moderated the age-related increase in TOTs during common noun retrieval.

Interestingly, no such age-related syllable frequency effect was observed during proper noun retrieval, as both high- and low-frequency initial syllables lead to an increase in TOTs among older adult participants (mean age = 68).

The TDH can account for the inordinate difficulty with retrieval of common nouns consisting of low-frequency initial syllables that older adults experienced. The age-related weakening of semantic-to-phonological connections (in this case at the level of the syllable) is exacerbated by the infrequent use of low-frequency syllables, thus demonstrating a cumulative effect detrimental to TOT rates. With regard to proper nouns, older adults experienced more

TOTs than younger adults for both low- and high-frequency first syllables, so it would appear that syllable frequency, in and of itself, does not influence the likelihood of phonological-level retrieval failures. Indeed, in Farrell’s 2012 study, older adults showed reduced TOTs for common nouns when provided with phonological primes, thereby supporting a breakdown at the phonological level of retrieval. By contrast, older adults did not benefit (i.e., there was no reduction in TOT incidence) from phonological primes during proper noun retrieval, which suggests that proper noun retrieval difficulties may not occur at the phonological level alone.

More specifically, these findings speak to the differential effect of psycholinguistic variables on common and proper nouns and likely structural differences. That is, as Farrell points out, there are numerous points of retrieval failure due to reduced transmission of activation, unique to the architecture of proper names 1) between the proper name phrase and the first name; 2) between the proper name phrase and the last name; 3) between the first name and its phonological components; 4) between the last name and its phonological components. Taken together, proper noun retrieval is negatively affected at both Step 1 and Step 2 of lexical retrieval with age, such that the beneficial effect of bottom-up activation from high-frequency first syllables observed in younger adults does not stave off TOTs for older adults.

To summarize, it is evident that lexical factors exert a differential influence on the semantic and/or phonological level during word production in healthy aging. These factors may

negatively act upon the production system independently or additively, and, in certain instances, may even interact in a way that protects successful retrieval. Adding an additional layer of complexity, given the distinction in the overall representation of common and proper nouns, these effects may vary by noun type. To date, no study has systematically investigated the influence of a variety of semantic- and phonologically-based psycholinguistic variables on the retrieval of common and proper nouns in older adults. The present dissertation additionally utilizes a methodological approach shown to be sensitive to retrieval difficulties at both stages of production (i.e., two-step approach), allowing for a fine-grained analysis of the locus of retrieval failures.

1.3. The Current Research

To review, the TDH, an interactive activation model, suggests that common nouns and proper nouns are differentially represented in the mental lexicon. The extent to which the representation of the two word types accounts for retrieval difficulties (i.e., TOTs) as a function of lexical factors known to interact with the semantic level, semantic-to-phonological level, and phonological level is underspecified. The present dissertation will examine the high- and low-

TOT-inducing capacity of a large set of common and proper nouns that are controlled for on a number of semantically- and phonologically-based features, all within a single experimental design. An examination of these features should allow for TOT researchers to better understand the independent contribution of specific lexical features to TOT occurrence in common and proper nouns in older adults. The overall purpose of the present investigation is to determine whether retrieval of common nouns and proper nouns is related to difficulties at the semantic and/or phonological level of processing. With this purpose in mind, this study has two specific aims:

• Specific Aim 1: To examine the role of semantically-based psycholinguistic features

(frequency and familiarity), noun type (common and proper), and cue modality (picture

and definition) on older adults’ retrieval difficulties at the semantic and phonological

levels of processing.

• Specific Aim 2: To examine the role of phonologically-based psycholinguistic features

(phonological neighborhood density, word length, and first-syllable frequency), noun

type, and cue modality on retrieval difficulties at the semantic and phonological levels of

processing.

1.4. Hypotheses

1.4.1. Specific Aim 1

• Hypothesis 1: The likelihood of a semantic stage failure or a phonological stage failure

occurring will increase as the frequency of the word decreases, and this will hold true for

both common and proper nouns.

Prior research supports the idea that target word frequency plays a role in retrieval of common nouns in both the first and second stage of processing such that low frequency words, referred to as “difficult” words, were more likely to result in a failure at both the semantic and phonological levels (Gollan and Brown, 2006). Additionally, there was an effect of word frequency on retrieval failures at stage one and stage two for proper nouns (Gianico-Relyea &

Altarriba, 2012), wherein low frequency words lead to significantly more naming failures at both stages compared to high frequency words. It should be noted that Gianico-Relyea & Altarriba

(2012) only employed Gollan and Brown’s (2006) set of calculations on the analysis of TOT occurrence and not on semantic-level failures. However, their data suggest that low frequency

words also led to more “don’t know” responses than high-frequency words, which implies an effect of frequency at the semantic level.

• Hypothesis 2: As familiarity with the target decreases, retrieval failures for common

nouns and proper nouns are more likely to occur at both the semantic and phonological

stages.

An additional psycholinguistic variable previously shown to influence semantic- and phonological-level retrieval difficulties is the familiarity of the target. Gollan and Brown (2006) suggest that common nouns that are both relatively low frequency and low familiarity, referred to as “difficult” targets, are more likely to result in failure at the first and second stage of retrieval as compared to “easy” targets (relatively higher frequency and higher familiarity). With respect to proper nouns, prior studies indicate that low familiarity words are more likely to lead to retrieval difficulties at both stages of retrieval compared to high familiarity targets (Hanley &

Chapman, 2008). Although these findings partially replicate those of previous studies, what is novel is that previous research has not determined the independent contribution of frequency and familiarity to the two stages of lexical processing for both noun types

1.4.2. Specific Aim 2

• Hypothesis 3: There will be no effect of word length on the semantic level for either noun

type. However, for both common and proper nouns, the likelihood of a phonological

stage failure occurring will increase as the length of the word (in phonemes) increases.

An additional goal of the present investigation is to determine whether retrieval failures at the semantic or phonological level are modulated by lexical characteristics known to influence the phonological stage of word retrieval, in particular, word length, first-syllable frequency, and neighborhood density. With respect to word length, no study has examined the effect of common

noun word length on the likelihood of a retrieval failure at the semantic level. One reason for this might be that word length is not believed to affect the semantic system. At the phonological level, by contrast, Hanley and Vandenberg (2010) found that word form retrieval failures were more likely for long common nouns (≥6 phonemes) than for short common nouns (≤4 phonemes). A similar pattern was observed in a study of proper noun retrieval such that longer words (3-word length) had a greater propensity for retrieval failure at the phonological level than shorter words (2-word length). There was no effect of word length at the semantic stage (Hanley

& Chapman, 2008). The effect of word length at the phonological level of retrieval can be described within a TDH framework: as length increases, the corresponding phonological nodes that need activation also increase, which creates additional opportunities for reduced transmission and failed retrieval.

• Hypothesis 4: In light of the paucity of literature on the effect of first-syllable frequency

on semantic stage retrieval performance, that effect will be specific to the phonological

level and that there will be no influence on retrieval at the semantic level for both noun

types. With regard to the phonological level, as first-syllable frequency increases,

common nouns will be less likely to result in a phonological stage failure. Conversely, an

increase in first-syllable frequency will increase the likelihood of a failure at the

phonological failure for proper nouns.

Prior research on healthy older adults examined the effects of first-syllable frequency on the phonological retrieval of both common and proper nouns (Farrell, 2012). The results indicated that, for common nouns, high frequency first syllable targets were less likely to result in failed phonological retrieval than low frequency targets. Interestingly, by contrast, proper nouns with high frequency first syllables increased the propensity for a TOT. The influence of

first-syllable frequency was previously thought to be relegated to the phonological stage of processing. However, Farrell and Abrams (2014) conducted a picture-word interference task on picture naming response times for common nouns and found that first-syllable frequency plays a role in lexical selection (i.e., semantic stage). This approach, however, may not be comparable to the experimental paradigms discussed thus far as Farrell and Abrams (2014) employed a task designed to increase competition through the presentation of a distractor word.

• Hypothesis 5: There will be no effect of phonological neighborhood density on either

semantic or phonological retrieval for common nouns

Neighborhood density is a lexical characteristic available only for common nouns (no corpus calculates this measure for proper nouns). The few studies that examined the effect of this feature on common noun retrieval have focused on its impact on TOT occurrence (Vitevitch

& Sommers, 2003). This is likely motivated by the fact that neighborhood density is considered a form-related word property (Newman & German, 2005). Failures at the phonological level have been found in the retrieval of words from sparse neighborhoods as compared to dense neighborhoods in a group of young adults (Harley & Bown, 1998). This effect was not observed in a separate study of healthy older adults (Vitevitch & Sommers, 2003). There was no main effect of neighborhood density on TOT likelihood; rather, an interaction was observed such that phonological level performance was reduced for words from a sparse neighborhood with low neighborhood frequency. One possibility is that words from dense neighborhoods experience age-related decrements in the transmission of priming that reduce their beneficial feedback activation from the phonological level to the lemma level. This would then lead to a similar effect of neighborhood density on TOT propensity regardless of whether the neighborhood is sparse or dense.

2. Methods and Materials

2.1. Participants

Participants were recruited from the tri-state area (New York, New Jersey, and Connecticut) through a number of methods including online advertisements and fliers placed in the community (e.g., senior centers). The total participant pool consisted of 52 older adults (M =

68.3, SD = 9.2) and consisted of 67% females and 33% males. Older adults were paid $12/hour for their time. See Table 1 for demographic characteristics. All participants were monolingual

English-speaking individuals with normal or corrected-to-normal hearing and vision. As well, they had no history of a learning disorder, cognitive impairment, psychiatric disorder, or traumatic brain injury. Participants were additionally screened for stroke history and other neurological impairments that might negatively impact performance. In addition, in order to gather more information about participants’ health and education status, all participants completed a general health and education background questionnaire. Descriptive statistics of participants’ age, years of education, and perceived overall health (on a 5-point scale: 5 is

“excellent” and 1 is “poor”) are outlined in Table 1. The study was approved by the City

University of New York’s Institutional Review Board.

Table 1

Demographic Characteristics

M (SD) Number of participants 52 Age (in years) 68 (9.2) Education (in years) 16 (2.3) Gender (female) 35 Health Rating (out of 5) 4.1 (0.6)

2.2. Materials

A total of 1,102 words were tested, consisting of 587 proper nouns and 515 common nouns. The lexical targets were obtained from prior TOT studies, including 454 common nouns and 95 proper nouns (Abrams, Trunk, & Margolin, 2007; Astell & Harley, 1996; Beattie & Coughlan,

1999; Biedermann, Ruh, Nickels, & Coltheart, 2008; Brennen, Baguley, Bright, & Bruce, 1990;

Brown & Nix, 1996; Burke et al., 1991; Farrell, 2012; Farrell & Abrams, 2014; Faust,

Dimitrovsky, & Davidi, 1997; Frick-Horbury, & Guttentag, Georgieff, Dominey, Michel, Marie-

Cardine, & Dalery, 1998; Gollan & Brown, 2006; Gollan & Silverberg, 2001; Hanley, 2011;

Harley & Bown, 1998; Jönsson, Tchekhova, Lönner, & Olsson, 2005; Juncos-Rabadán et al.,

2010; Kohn et al., 1987; May & Clayton, 1973; Meyer & Bock, 1992; Perfect & Hanley, 1992;

Pine, Bird, & Kirk, 2007; Pyers, Gollan, & Emmorey, 2009; Vigliocco, Vinson, Martin, &

Garrett, 1999; Yaniv & Meyer, 1997), and from the Boston Naming Test (Kaplan, Goodglass, &

Weintraub, 1983). As well, additional common (n = 61) and proper nouns (n = 492) were compiled by the author in order to add more words to the stimulus set that have the potential to induce TOTs. Proper nouns consisted of only famous people; no places (see Appendix A for the stimulus set and associated psycholinguistic variables). The additional proper noun targets were composed of a variety of occupational categories that ranged from the early 20th century through present day: politicians (Herbert Hoover), historical figures (Rosa Parks), artists (Salvador

Dali), writers (J.D. Salinger), directors (Quentin Tarantino), musicians (Beyonce), actors

(Shirley Temple), celebrities (Twiggy) and athletes (Shaquille O’Neal). These targets were obtained by conducting systematic google searches (“politicians of the 1940s”) and by accessing online databases (oscars.org, thefamouspeople.com).

Picture cues were color photographs that do not contain semantic clues (e.g., no baseball uniform for Babe Ruth; no farm for silo) that might aid retrieval. All photographs were obtained through internet searches. For proper nouns, the image was either a “head shot” or a “three- quarter” view of the individual. Image dimensions were resized to a pixel height of 318 and width of 272, which relates to a pixels per inch of 3.778 inches wide by 3.317 inches high. For both proper noun and common nouns, definition cues taken from prior studies were adapted as needed to reflect colloquial language use. For the additional targets created by the author, common noun definitions were selected from several online dictionaries (e.g., Merriam Webster

Dictionary, Oxford English Dictionary, etc.). Newly-generated proper noun target definitions were constructed based on biographical information obtained through internet searches (e.g., the films that an actor is most famous for were retrieved from the online movie database IMDb). In line with Farrell (2012), all proper noun definitions contained at least three biographical details as participants may have been exposed to the target through different frames of reference.

Consistent with prior TOT research, the definition cue is a trivia-like question designed to elicit a single, appropriate target. For common noun targets, the question required participants to retrieve the word that best reflects the definition (e.g., What is the name for an ancient Egyptian stone figure having a lion’s body and a human or animal head, especially the huge statue near the

Pyramids at Giza? Target = sphinx). For proper noun targets, participants retrieved the name of a famous person from the biographical content presented in the definition (e.g., What is the name for the American director and screenwriter whose films are noted for their stylized violence and razor-sharp dialogue, and among which we can find Reservoir Dogs, Pulp Fiction, Kill Bill, and

Django Unchained? Target = Quentin Tarantino).

With respect to the linguistic features of the targets, several psycholinguistic databases were utilized to extract relevant content. Word length in syllables, letters, and phonemes was calculated using the English Lexicon Project (ELP) database (Balota et al., 2007). Common noun and proper noun targets ranged from 1-5 syllables in length (M = 2.8; M = 2.3, respectively). For common nouns, the letter length ranged from 4-15 (M = 7.4) and length in phonemes ranged from 3-13 (M = 6.32). The letter length of proper nouns spanned 4-20 (M = 11.6) and the number of phonemes ranged from 4-18 (M = 9.9). Three corpora-derived psycholinguistic properties pertained only to common nouns: frequency, familiarity, and phonotactic probability, as such measures are unavailable in English lexicon databases for proper nouns. Phonological neighborhood density (the quantity of words that differ from a target by an added, deleted, or substituted phoneme) was collected from the ELP database (Balota et al., 2007). Value ranged from 1 to 59 (M = 11.47). Phonological neighborhood density values were available for 296

(57%) of the common nouns. The linguistic property of familiarity (how well known a word is) was derived from the MRC Psycholinguistic Database (Coltheart, 1981), and values spanned from 168 to 613 (M = 458.37). This measure will be referred to as ‘MRC familiarity” in order to distinguish it from self-rated familiarity values. MRC familiarity values were available for 258

(50%) of the common nouns targets. Finally, Zipf frequency was constructed from the

SUBTLEX-US corpus (Van Heuven et al., 2014). This measure of frequency is standardized and ranges from 1-7. Low frequency words fall between 1 and 3 on the scale and high frequency words from 4 to 7. The average Zipf value was 3.29, ranging from 1.59 to 6.48. This measure is henceforth referred to as ‘Zipf frequency.’ The Zipf frequency value was available for 510 (99%) of the common noun targets.

First-syllable frequency for all targets was calculated using the CELEX-2 database

(Baayen, Piepenbrock, & Gulikers, 1995), which is the only normative psycholinguistic database in English that allows for measures of syllable frequency (Farrell & Abrams, 2011; Farrell,

2012). The database predominately reflects common noun usage; however, there are instances of proper nouns in the corpus. For both common and proper nouns, first-syllable frequency was calculated by summing the frequency of the first syllable of the first word. As an example, the onset syllable in Albert Einstein is [æl], which has a combined frequency per million of 87.

Taking an example from a common noun in the stimulus set, the first syllable [ræ] in radish occurs in words that have a summed frequency of 379 instances per million. Given that the word frequencies are based on CELEX-2’s corpus of 17.9 million words, the summed frequency was then divided by 17.9 to normalize the measure to reflect frequency per million (Macizo & Van

Petten, 2007). CELEX-2’s syllable bounds were utilized to extract the onset syllable of each word in the stimulus set, and an overall count was performed on instances of a given syllable out of all nominal lemma forms in the corpus (common and proper nouns).

Given the large stimulus set (1,102 items), the stimuli were randomly assigned to sublists in order to create a manageable set of targets for retrieval and subsequent rating. The total number of lexical items was divided into five lists. List one was composed of 220 targets (117 proper nouns and 103 common nouns); list two of 221 targets (118 proper nouns and 103 common nouns); list three of 224 targets (120 proper nouns and 104 common nouns); list four of

217 targets (114 proper nouns and 103 common nouns); list five of 220 targets (118 proper nouns and 102 common nouns). Each of the five lists was further subdivided into seven1 additional sub-lists, and lexical targets were randomly assigned within those lists. In addition, all

1 The division of each of the five lists into seven sublists was done to create a manageable testing set for participants and to accommodate the processing limitations of Google Forms.

lexical targets were cued from both a picture and definition; however, no participant saw both the picture cue and the definition cue of the same lexical target. For lists one through five, half of the lexical targets were presented as definitions and half as pictures. These lists and their sub-lists are referred to as stimulus set A. Cue type was counterbalanced such that another set of lists, referred to as Stimulus Set B, was created with lexical targets being presented in the opposite cueing modality from Stimulus Set A. For example, if Barbara Streisand was presented as a picture cue in one of the lists in Stimulus Set A, it was presented as a definition cue in one of the lists in Stimulus Set B. It should be noted that lexical targets were randomly assigned to Stimulus

Set B and thus did not follow the same order of presentation, within a list, as Stimulus Set A. See

Appendix B for a review of the total number of common nouns and proper nouns assigned to each sub-list in Stimulus Sets A and B. On average, participants attempted retrieval of 90 words

(three of the sublists just described), consisting of approximately 45 proper nouns and 45 common nouns from approximately a 50/50 distribution of picture and definition cues in each list.

2.3. Procedure

The experiment contained two parts: 1) a naming task in which participants were asked to retrieve the target items from either a picture or definition cue, and; 2) a rating component in which participants made lexical judgments about the target on three scales – how frequently they have used the word in their lifetime, how familiar they are with the word, as well as the clarity of the picture or definition cue. A member of the research team, either the author or a trained research assistant, was present throughout the experiment. Before the study commenced, participants were verbally provided with a definition of key terms in colloquial language (i.e., the meaning of TOT, frequency, and familiarity), instructions as to the workflow of the overall

experiment, and the experimenter’s expectations for each aspect of the task (see Appendix C).

The experimenter offered guidance and assistance in the first few trials as the participant became familiar with the protocol. This assistance was limited to the how-to’s of the task.

The lexical retrieval task was administered through Google Forms. Participants were instructed to provide a typed response as quickly as possible in order to minimize long testing sessions and to reduce instances in which a TOT may be experienced but participants wait to resolve it before responding. The ability to return to a previous answer was disabled in the survey. At the beginning of each trial, participants were presented with a definition or picture cue corresponding to a common noun or proper noun target. A short-answer space was provided for participants to enter a response. If the target word was known, the participant typed in the first and last name (for proper nouns) or the target word (for common nouns). Misspellings did not impact scoring so as long as the typed entry was unambiguous. In cases where the participant could not come up with the word, then “don’t know” or “DK” was entered. If the cue elicited a

TOT state, then the participant typed “tip of the tongue” or “TOT.” Following one of these three response types (retrieval attempt, DK, or TOT), the next screen provided the correct answer, and participants were asked through a dichotomous (yes/no) question if they indeed experienced a

TOT for the target word2. This step differentiated between positive TOTs (i.e., TOTs for the target item) and negative TOTs (i.e., off-target TOTs). For the purpose of this study, only positive TOTs will be included in analyses.

Rating scales followed the naming task in order to capture participants’ judgments of how frequently they have used the target word over their lifetime and how familiar they are with

2 The version of Google Forms used at the time of testing did not allow for conditional responses. Therefore, all participants were required to answer whether a TOT was experienced regardless of the response type originally indicated (i.e., retrieval attempt, DK, or TOT).

it. See Figures 4 through 7. In order to discourage participants from over-responding on the rating scales that they have never used a word and are unfamiliar with it, participants were told that the words and names were taken from a variety of prior studies as well as generated by the present author in order to understand whether the words are representative of the present generation of older adults’ vocabulary. The participants were then of the understanding that they were helping to develop a stimulus set reflective of older adults’ rich language knowledge, rather than being critically evaluated for what they did or did not know. Participants rated the frequency of a target word on a 5-point scale, with 1 representing “never” and 5 “very frequent.”

Familiarity was evaluated on the same 5-point scale with 1indicating “completely unfamiliar” and 5 “very familiar.” Participants were provided with a printout of the 5-point Likert scale with explanations for each point on the scale, and this was created for both common noun frequency and familiarity ratings as well as and proper noun frequency and familiarity ratings. These scales were modified from the familiarity scale of another TOT study (Farrell, 2012). The target word and the cue (picture or definition) were provided at the top of the screen for reference. Once a selection for each scale was made, participants could advance to the next trial by clicking “next.”

Figure 4

Frequency Rating Scale for Common Nouns

1 2 3 4 5 Never Infrequently Somewhat Frequently Very Frequently Frequently

I have never I have used this I use this word I occasionally I often use this used this word word a handful once in a while use this word word in my day- when speaking of times in in speaking or when speaking to-day life either or writing at any speaking or writing but not or writing day- in speaking or point in my life. writing in my day-to-day. to-day. writing. lifetime.

Figure 5

Frequency Rating Scale for Proper Nouns

1 2 3 4 5 Never Infrequently Somewhat Frequently Very Frequently Frequently

I have never I have referred I refer to this I occasionally I often refer to referred to this to this person a person once in a refer to this this person in person when handful of times while in person when my day-to-day speaking or in speaking or speaking or speaking or life either in writing at any writing in my writing, but not writing day-to- speaking or point in my life. lifetime. day-to-day. day. writing.

Figure 6

Familiarity Rating Scale for Common Nouns

1 2 3 4 5

Completely Somewhat Vaguely Familiar Very Familiar Unfamiliar Unfamiliar Familiar I have not heard I have heard the I have heard the I hear and use I hear and use this word and word but could word and the word the word could not define not define it or produced it occasionally. I regularly and it or describe use it in a before. I could could use the could provide a anything about it sentence not give a word correctly in clear definition definition or use a sentence and of the word it in a sentence, could easily but I may be define it able to select the correct definition if given choices

Note. Figure adapted from Farrell (2012).

Figure 7

Familiarity Rating Scale for Proper Nouns

1 2 3 4 5

Completely Somewhat Vaguely Familiar Very Familiar Unfamiliar Unfamiliar Familiar I have never I think I have I have heard or I hear and talk I talk about this heard of this heard this name talked about this about this person person often and person and could before, but I am person before, occasionally. I could tell you not tell you a fairly sure that I but it would be could probably many facts about single fact about have never difficult to tell tell you a couple him or her him or her talked about this you more than a of facts about person. I could single fact about him or her not tell you him or her anything about him/her with any confidence

Note. Figure adapted from Farrell (2012).

2.4. Statistical Analysis

In order to understand the influence of semantically-based psycholinguistic features (i.e., frequency and familiarity) and phonologically-based psycholinguistic features (i.e., phonological neighborhood density, letter length, and first-syllable frequency) on the retrieval of common and proper nouns, two error types were examined: semantic stage failure (step 1) and phonological stage failure (step 2). As outlined earlier, both error types are implicated in specific lexical retrieval stages according to two-stage models of retrieval (Gollan & Brown, 2006; Juncos-

Rabadán et al., 2010; Kittredge et al., 2008). Analyses were conducted on a trial level; every word in the stimulus set had approximately four trials (i.e., four participants saw a given target in their respective testing sets), and thus each presentation of a word was considered a trial.

Participant-level analyses could not be conducted as the majority of participants received different overall testing sets.

The calculations proposed by Gollan & Brown (2006) were adapted in order to align with trial-level analysis. More specifically, proportions could not be utilized because the calculation takes into account the quantity of the various response types an individual can make across all trials.

According to Gollan & Brown (2006), semantic stage success is reflected in both GOT and +TOT responses and phonological stage success in GOTs alone. Therefore, in the present study, semantic stage failure (Step One) was indicated by the presence of a Not GOT response and semantic stage success by a GOT or +TOT. A Not GOT response refers to a situation in which a word is known but not retrieved (i.e., failed semantic access). Phonological stage failure was determined by whether a TOT occurred on the trial. Importantly, if a trial was coded as a semantic stage failure then success or failure at the phonological level could not occur since a

Not GOT precludes phonological access and the trial was not included in analyses of phonological stage failures. Additionally, any trial that resulted in an NK response was not included in the semantic and phonological stage failure analyses because a word that is not known could not have had an opportunity for success at either stage of retrieval. See Figure 8.

A series of binary logistic regression analyses were employed to examine the effect of the semantically- and phonologically-based psycholinguistic predictors on retrieval failures at the semantic and phonological level for common nouns and proper nouns, while accounting for known noun type effects and possible cue modality effects. Binary logistic regressions are employed when the dependent variable is categorical (dichotomous), and the statistic allows for a mix of categorical and continuous predictor variables, which need not be normally distributed

(e.g., Peng et al., 2002; Redmond et al., 2011). Assumption testing for the binary logistic regression was conducted in order to improve the accuracy of the models (i.e., independence of

observations, linearity, multicollinearity, and sample size). The first assumption relates to the independence of the observations – the dependent variable as well as the nominal independent variables must be mutually exclusive. In the present study, this assumption is met since, for any given trial, the dependent variable (e.g., semantic stage retrieval) is coded as either a ‘failure’ or a ‘success,’ never both. This also holds true for the nominal independent variables (e.g., cue modality in a single trial can only refer to a picture or a definition). In addition, trials reflect each presentation of a word across all participants, which means that data from each participant appears multiple times in the total trials analyzed (consistent with the number of words in a given participant’s data set). Importantly, each trial is unique—the response to one target did not play a role (positive or negative) in subsequent word presentations; however, measurements from the same participant cannot be assumed to be uncorrelated, which may reduce the statistical independence.

Regarding the linearity assumption, a Box-Tidwell test was conducted on the continuous predictors in each model (word length in phonemes, first-syllable frequency, Zipf frequency,

MRC familiarity, and phonological neighborhood density) (Hosmer & Lemeshow, 1989). Within the models, significant interaction terms are noted as having violated the linearity of the logit assumption. Multicollinearity was assessed among the independent variables via correlation matrix (see Table 2). Given the mild-to-moderate associations among variables, Variance

Inflation Factors (VIFs) were calculated using a partial linear model for all eight independent variables (cue modality, subjective frequency, subjective familiarity, length-in-phonemes, first- syllable frequency, Zipf frequency, MRC familiarity, and phonological neighborhood density), none of which were over 5 (a standard conservative threshold for multicollinearity; Menard,

2001). These diagnostic measures indicate that multicollinearity is not a problem for the dataset.

See Tables 2 and 3.

Figure 8

Determination of Success or Failure at Two Stages of Lexical Processing

Note. ‘Never Knew’ items were excluded from analysis. ‘Not GOT’ indicates semantic stage retrieval failure, so phonological retrieval stage analysis was not conducted for that trial. For trials resulting in semantic retrieval success (‘GOT’ or ‘+TOT’), then the phonological retrieval stage was analyzed for success or failure. A ‘+TOT’ is coded as a phonological stage retrieval failure.

Table 2

Correlation matrix of independent variables

Sub. Sub. Length- MRC Zipf Phono. freq. fam. in-Phon. Fam. Freq. ND Sub. fam. Pearson Correlation .546** Sig. (2-tailed) .000 N 4370 Length-in- Pearson Correlation -.168** -.166** phonemes Sig. (2-tailed) .000 .000 N 4375 4370 MRC Pearson Correlation .431** .349** -.205** Fam. Sig. (2-tailed) .000 .000 .000 N 1027 1025 1028 Zipf Pearson Correlation .367** .342** -.341** .710** Freq. Sig. (2-tailed) .000 .000 .000 .000 N 2028 2024 2036 1028 Phono. ND Pearson Correlation .043 .089** -.636** .150** .224** Sig. (2-tailed) .139 .002 .000 .000 .000 N 1178 1177 1182 737 1182 First-syl. Pearson Correlation .020 -.017 .076** .012 -.114** -.218** freq. Sig. (2-tailed) .186 .266 .000 .708 .000 .000 N 4245 4240 4272 1028 2028 1182

Note. Sub. freq. = subjective frequency; Sub. Fam. = subjective familiarity, Length- in-phon. = length-in-phonemes; MRC Fam. = MRC familiarity; Zipf Freq. = Zipf frequency; Phono. ND = phonological neighborhood density; First-syl. freq. = first-syllable frequency

** Correlation is significant at the 0.01 level (2-tailed)

Table 3

Collinearity values for independent variables

Collinearity Statistics Model Tolerance VIF 1 Modality .999 1.001 Sub. frequency .788 1.269 Sub. familiarity .838 1.194 Length-in-phonemes .551 1.814 First-syllable frequency .910 1.098 MRC Familiarity .512 1.952 Zipf Frequency .539 1.856 Phonological ND .560 1.787

Note. Sub. frequency = subjective frequency; Sub. familiarity = subjective familiarity, Length in phon. = length in phonemes; Phonological ND = phonological neighborhood density

Power analyses were conducted using the Power Analysis of Univariate Linear

Regression to determine the minimum required sample size. The analysis employed a two-tailed test, taking into account the eight independent variables (i.e., the most predictors entered into a single model), using a small effect size of .1 and α = .05. The results indicate that a total sample of 1,495 cases are minimally required to achieve a power of .80 for logistic regression analyses

(see Table 4). The overall sample size is not a problem since the present study comprises 4,375 trials. Models that fall short of the minimum sample are noted and interpreted with caution.

Table 4

Power analysis

Actual Predictors Test Assumptions N Powerb Total Test Power Partialc Sig. Type III F-testa 1495 .800 8 8 .8 .1 .05 a. Intercept term is included, b. Predictors are assumed to be fixed, c. Multiple partial correlation coefficient.

A total of six binomial logistic regressions were conducted. The first pair of regressions examined the contributions of noun type (common noun vs. proper noun), cue modality (picture vs. definition cue), subjective frequency (SR frequency), subjective familiarity (SR familiarity), length-in-phonemes, and first-syllable frequency to semantic stage access (regression one) and phonological stage access (regression two). For both regressions, outcome measures were coded such that ‘0’ meant that the stage was accessed successfully and ‘1’ indicated that access to the stage failed. These regressions included both common and proper nouns in order to examine overall effects of the semantic and phonological variables in the stimulus set as well as to look at the effect of noun type. The second set of regressions were run on common nouns only and evaluated the effect of all the explanatory variables from the first set of regressions as well as variables for which data exist only for common nouns, Zipf frequency, MRC familiarity, and phonological neighborhood density, on semantic and phonological access. Finally, the third set of regressions were conducted on proper name stimuli with the same set of predictors and outcome variables as the first set of regressions. The models for proper nouns alone were designed to examine the effects of the psycholinguistic properties on retrieval performance at the semantic and phonological stages for proper nouns, specifically.

Descriptive statistics were calculated for all variables included in the analyses. There were 4,404 total trials, of which 29 contained missing information, due to either user error or to survey platform error (i.e., Google Forms). Therefore, 4,375 total trials were included for analysis. Trials for which the target was never known (13.9% of trials, 612 out of 4,375 trials) were excluded from all analyses. As well, trials for which a semantic stage failure occurred were removed from phonological stage analyses (17.4% of trials, 759 out of 4,375 trials). Significance for all analyses was set at p < .05. See Table 5 for the distribution of response types across trials.

Analyses of the relationship between demographic factors of age, education, and gender on lexical retrieval performance (semantic and phonological stage errors) were conducted. This set of analyses was run at the participant level, rather than the trial level. Specifically, the percentage of semantic stage and phonological stage errors are calculated for each participant using the calculations specified by Gollan and Brown (2006). The association between age, education, and percentage of semantic stage and phonological stage retrieval errors are analyzed using Pearson correlations. An effect of gender was examined through an independent samples t- test.

All data were analyzed in IBM SPSS Statistics Version 25 (IBM Corp., 2017).

Table 5

Response types across trials

Response Type N % (of 4,375 trials) NK 612 13.9% GOT 2188 50% Not GOT (semantic stage failure) 759 17.4% TOT (phonological stage failure) 816 18.7%

Note. NK = never knew, GOT = correct retrieval, Not GOT = word is known but not retrieved,

TOT = tip-of-the-tongue

3. Results

3.1. Semantic stage lexical retrieval

In sections 3.1.1-3.1.3, Tables 6-9 present the regression results for models examining the effect of predictors of semantic stage failure given common and proper nouns, common and proper nouns (with noun type*cue modality interaction term added), common nouns alone, and proper nouns alone, respectively. In section 3.2.3, Tables 14-16 summarize the overall results for models from the full dataset, common nouns alone, and proper nouns alone, respectively.

3.1.1. Full dataset (common and proper nouns)

A binary logistic regression with cue modality (picture vs. definition), noun type (common vs. proper), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency predictors of semantic stage failure yielded an overall significant result, χ2 (8, N =

3,652) = 285.39, p < .001, Nagelkerke’s R2 = .119. The results are summarized in Table 6. All models in this section include both common and proper nouns considered together.

Table 6

Regression results for cue modality, noun type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of semantic stage failure

95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modalitya .041 .086 .220 1 .639 1.041 .879 1.234 Noun typeb -.058 .128 .208 1 .648 .943 .734 1.212 Sub. frequency -.334 .041 65.194 1 <.001 .716 .661 .777 Sub. familiarity -.492 .048 103.080 1 <.001 .611 .556 .672 Length in phon. -.333 .192 3.009 1 .083 .717 .492 1.044 First-Syl. Freq. -.008 .013 .372 1 .542 .992 .967 1.018 Constant 2.452 .563 18.977 1 .000 11.608

Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency;

Sub. familiarity = subjective familiarity; Length in phon. = length in phonemes; First-syl. freq. = first-syllable frequency aDefinition = 0, bProper noun = 0

3.1.1.1. Noun type and cue modality

The binomial logistic regression revealed that neither cue modality (B = .041, SE =.086, Wald =

.220, p = .639) nor noun type (B = -.058, SE =.086, Wald = .128, p = .648) were significant predictors of semantic stage failure when accounting for the effect of the other predictors. To examine the influence of the interaction of noun type and cue modality as a predictor of semantic stage failure likelihood, the same binomial logistic regression was run with the addition of the interaction term. Noun type*cue modality was not a significant predictor of semantic stage failure (B = -.333, SE =.192, Wald = 3.009, p = .083). The results are summarized in Table 7.

Table 7

Regression results for cue modality, noun type, noun type*cue modality, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of semantic stage failure

B SE Wald df Sig. OR 95% CI for OR Lower Upper Cue modalitya .919 .957 .992 1 .337 2.506 .384 16.347 Noun typeb 1.615 1.821 .787 1 .375 5.027 .142 178.219 SR frequency -.335 .041 65.533 1 <.001 .715 .660 .776 SR familiarity -.492 .048 102.952 1 <.001 .612 .556 .672 Noun type*cue .159 .173 .849 .357 1.173 .836 1.646 modality Length in phon. -.331 .192 2.986 1 .084 .718 .493 1.045 First-syl. Freq. -.008 .013 .366 1 .545 .992 .967 1.018 Constant -8.030 11.390 .497 1 .481 .000

Note. χ2 (9, N = 3,652) = 286.243, p < .001, Nagelkerke’s R2 = .119, CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency; Sub. familiarity = subjective familiarity, Length in phon. = length in phonemes; First-syl. freq. = first- syllable frequency. aDefinition = 0, bProper noun = 0

3.1.1.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed a significant decrease in the log odds of having a semantic stage failure per unit increase in subjective frequency (B = -.334, SE =.041, Wald =

65.194, p < .001). This suggests that a semantic stage failure is 28% less likely to occur when frequency increases than when it decreases3. Similarly, subjective familiarity was found to significantly decrease the log odds of experiencing a retrieval failure at the semantic stage (B = -

3 When the odds ratio is less than 1, the % likelihood of the outcome (XX%) is calculated using the following formula: (XX% = 1 - odds ratio). An odds ratio less than one corresponds to a % decrease in the odds of an outcome.

.492, SE =.048, Wald = 103.080, p < .001). This finding indicates that a semantic stage failure is

39% less likely to occur when familiarity is high. Taken together, greater frequency and familiarity ratings have an increased likelihood of successful semantic access.

3.1.1.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that neither word-length-in-phonemes (B = -.333, SE

=.192, Wald = 3.009, p = .083) nor first-syllable frequency (B = -.008, SE =.013, Wald = .372, p

= .542) were significant predictors of semantic stage failure when accounting for the effect of the other predictors.

Table 14 summarizes the overall results for models of semantic stage failure from the full dataset.

3.1.2. Analysis of Common Nouns

All models in this section include only common noun trials. The results of the models presented in this section should be interpreted with caution, as the minimum sample size for detecting a small effect was not met (see Button et al., 2013). A binary logistic regression with semantic stage failure regressed onto cue modality (picture vs. definition), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as well as three predictors that are relevant to the common noun data set alone (MRC familiarity, Zipf frequency, and phonological neighborhood density) yielded an overall significant result of the model, χ2 (13, N = 712) =

46.031, p < .001, Nagelkerke’s R2 = .104. The results are summarized in Table 8.

3.1.2.1. Cue modality

The binomial logistic regression revealed that cue modality (B = .047, SE =.206, Wald = .051, p

= .822) was not a significant predictor of semantic stage failure when accounting for the effect of the other predictors.

Table 8

95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modalitya .047 .206 .051 1 .822 1.048 .699 1.570 Sub. frequency -.117 .096 1.469 1 .226 .890 .737 1.075 Sub. familiarity -.573 .140 16.704 1 <.001 .564 .428 .742 Length in phon. -1.211 1.229 .972 1 .324 .298 .027 3.308 First-syl. freq. -.095 .068 1.939 1 .164 .910 .796 1.039 MRC familiarity .036 .101 .128 1 .721 1.037 .851 1.263 Zipf frequency -.966 3.052 .100 1 .752 .381 .001 150.723 Phonological ND .060 .081 .544 1 .461 1.062 .905 1.245 Constant 3.056 6.240 .240 1 .624 21.245

Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency;

Sub. familiarity = subjective familiarity; Length in phon. = length in phonemes; First-syl. freq. = first-syllable frequency, ND = neighborhood density aDefinition = 0

3.1.2.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed that subjective frequency was not a significant predictor of semantic stage retrieval failure (B = -.117, SE =.096, Wald = 1.469, p = .226). Subjective familiarity was found to significantly decrease the log odds of experiencing a retrieval failure at the semantic stage (B = -.573, SE =.140, Wald = 16.704, p < .001). This finding indicates that a semantic stage failure is 44% less likely to occur when subjective familiarity rating increases than when it decreases. Taken together, common nouns rated by participants as high familiarity have an increased likelihood of successful semantic access.

3.1.2.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that length-in-phonemes (B = -1.221, SE =1.229, Wald

= .972, p = .324), first-syllable frequency (B = -.095 , SE = .068, Wald = 1.939, p = .164), MRC familiarity (B = .036 , SE = .101, Wald = .128 , p = .721 ), Zipf frequency (B = -.966 , SE =

3.052, Wald = .100 , p = .752), and phonological neighborhood density (B = .060 , SE =.081 ,

Wald = .544 , p = .461 ) were not significant predictors of semantic stage failure during common noun retrieval when accounting for the effect of the other predictors.

Table 15 summarizes the overall results for models of semantic stage failure for common nouns.

3.1.3. Analysis of Proper Nouns

All models in this section include only proper noun trials. A binary logistic regression with semantic stage failure regressed onto cue modality (picture vs. definition), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency yielded an overall significant result of the overall model, χ2 (7, N = 1730) = 166.885, p < .001, Nagelkerke’s R2 =

.141. The results are summarized in Table 9.

3.1.3.1. Noun type and cue modality

The binomial logistic regression indicated that cue modality (B = .041, SE =.086, Wald = .220, p

= .639) was not a significant predictor of semantic stage failure for proper noun retrieval when accounting for the effect of the other predictors.

3.1.3.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed that subjective frequency significantly reduced the log odds of experiencing a retrieval failure at the semantic stage per unit increase in subjective frequency (B = -.489, SE =.064, Wald = 58.99, p < .001). This suggests that a semantic stage

failure is 39% less likely to occur when frequency increases than when it decreases. Subjective familiarity was similarly found to significantly decrease the log odds of experiencing a retrieval failure at the semantic stage per unit increase in subjective familiarity (B = -.412, SE =.063,

Wald = 42.581, p < .001). This finding indicates that a semantic stage failure is 34% less likely to occur as familiarity increases. Taken together, proper nouns rated by participants as high frequency or high familiarity have an increased likelihood of successful semantic access.

Table 9

Regression results for cue modality, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of proper noun semantic stage failure

95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modality -.053 .122 .187 1 .665 .949 .747 1.205 Sub. frequency -.489 .064 58.991 1 <.001 .613 .541 .695 Sub. familiarity -.412 .063 42.581 1 <.001 .663 .586 .750 Length in phon. -.521 .492 1.122 1 .289 .594 .226 1.558 First-syl. freq. .012 .019 .423 1 .516 1.012 .975 1.051 Constant 3.035 1.519 3.993 1 .046 20.802

Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency; Sub. familiarity = subjective familiarity, Length in phon. = length in phonemes; First-syl. freq. = first-syllable frequency, ND = neighborhood density aDefinition = 0

3.1.3.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that neither length-in-phonemes (B = -.521, SE =.492,

Wald = 1.122, p = .289) nor first-syllable frequency (B = .012 , SE = .019, Wald = .423, p =

.516), were significant predictors of semantic stage failure during proper noun retrieval when accounting for the effect of the other predictors.

Table 16 summarizes the overall results for models of semantic stage failure from proper nouns alone.

3.2. Phonological stage lexical retrieval

In sections 3.2.1-3.2.3, Tables 10-13 present the regression results for models examining the effect of predictors of phonological stage failure given common and proper nouns, common and proper nouns (with noun type*cue modality interaction term added), common nouns alone, and proper nouns alone, respectively. In section 3.2.3, Tables 14-16 summarize the overall results for models from the full dataset, common nouns alone, and proper nouns alone, respectively.

3.2.1. Overall analysis (common and proper nouns)

All models in this section include both common and proper nouns. A binary logistic regression with cue modality (picture vs. definition), noun type (common vs. proper), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency predictors of phonological stage failure yielded an overall significant result of the model, χ2 (8, N = 2,918) =

527.52, p < .001, Nagelkerke’s R2 = .242. The results are summarized in Table 10.

Table 10

Regression results for cue modality, noun type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of phonological stage failure

95% CI for OR B SE Wald df Sig. OR Lower Upper SCue modality -.357 .093 14.596 1 <.001 .700 .583 .841 tNoun type -1.129 .138 66.488 1 <.001 .323 .247 .424 eSub. frequency -.267 .042 39.784 1 <.001 .766 .705 .832 pSub. familiarity -.772 .065 141.684 1 <.001 .462 .407 .525 1Length in phon. .386 .231 2.801 1 .094 1.472 .936 2.314 a First -syl. frequency .017 .014 1.379 1 .240 1.017 .989 1.046 Constant 2.761 .708 15.227 1 .000 15.815

Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency; Sub. familiarity = subjective familiarity; Length in phon. = length in phonemes; First-syl. freq. = first-syllable frequency aDefinition = 0, bProper noun = 0

3.2.1.1. Noun type and cue modality

The binomial logistic regression revealed a significant decrease in the log odds of a phonological stage failure when the cue modality is a picture cue as compared to a definition cue (B = -.357,

SE =.093, Wald = 14.596, p < .001). More specifically, this indicates that the odds of having a retrieval failure at the phonological level is 30% less likely for picture cues than for definition cues. A significant decrease in the log odds of a phonological stage failure was also found for noun type such that common nouns were 68% less likely to lead to retrieval difficulty at the phonological level in comparison to proper nouns (B = -1.129, SE =.138, Wald = 66.488, p <

.001).

To further explore the influence of the interaction of noun type and cue modality as a predictor of phonological stage failure likelihood, the same binomial logistic regression was run with the addition of the interaction term. Noun type*cue modality was a significant predictor of phonological stage failure (B = -.542, SE =.194, Wald = 7.779, p < .01). The results are summarized in Table 11. The findings suggest that when the noun type is a common noun and the cue modality is a picture, the odds ratio of a phonological retrieval failure is reduced by 42% as compared to proper nouns with definition cues.

3.2.1.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed a significant decrease in the log odds of having a phonological stage failure per unit increase in subjective frequency (B = -.267, SE =.042, Wald =

39.78, p < .001). This suggests that a phonological stage failure is 23% less likely to occur when frequency increases than when it decreases. Similarly, subjective familiarity was found to significantly decrease the log odds of experiencing a retrieval failure at the phonological stage (B

= -.772, SE =.065, Wald = 141.684, p < .001). This finding indicates that a phonological stage failure is 54% less likely to occur when familiarity increases than when it decreases. Taken together, words rated by participants as high frequency or high familiarity have an increased likelihood of successful phonological access compared to words of lower frequency and familiarity.

Table 11

Regression results for cue modality, noun type, noun type*cue type, subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as predictors of phonological stage failure

95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modalitya -3.402 1.097 9.613 1 <.01 .033 .004 .286 Noun typeb -6.851 2.059 11.070 1 <.01 .001 .000 .060 Noun type*Cue modality -.542 .194 7.779 1 <.01 .582 .397 .851 SR frequency -.264 .042 39.027 1 <.001 .768 .707 .834 SR familiarity -.773 .065 141.25 1 <.001 .462 .406 .524 8 Length in phon. .384 .232 2.754 1 .097 1.469 .933 2.312 First-syl. frequency .017 .014 1.388 1 .239 1.017 .989 1.046 Constant 38.433 12.813 8.997 1 .003 49130550849

296400.000

Note. χ2 (9, N = 2,918) = 535.385, p < .001, Nagelkerke’s R2 = .245, CI = confidence interval,

OR = odds ratio; Sub. frequency = subjective frequency; Sub. familiarity = subjective familiarity; Length in phon. = length in phonemes; First-syl. freq. = first-syllable frequency aDefinition = 0, bProper noun = 0

3.2.1.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that neither length-in-phonemes (B = .386, SE =.231,

Wald = 2.801, p = .094) nor first-syllable frequency (B = .017, SE =.014, Wald = 1.379, p =

.240) were significant predictors of phonological stage failure when accounting for the effect of the other predictors, though length-in-phonemes approached significance.

The overall results for models of phonological stage failure from the full dataset are presented in Table 14.

3.2.2. Analysis of Common Nouns

All models in this section only include common noun trials. The results of the models presented in this section should be interpreted with caution, as the minimum sample size for detecting a small effect was not met (see Button et al., 2013). A binary logistic regression with semantic stage failure regressed onto cue modality (picture vs. definition), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency as well as three predictors that are relevant to the common noun data set alone, MRC familiarity, Zipf frequency, and phonological neighborhood density, yielded an overall significant result, χ2 (13, N = 589) = 55.614, p < .001,

Nagelkerke’s R2 = .177. The results are summarized in Table 12.

Table 12

Regression results for cue modality, subjective frequency, subjective familiarity, length-in- phonemes, first-syllable frequency, MRC familiarity, Zipf frequency, and phonological neighborhood density as predictors of common noun phonological stage failure 95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modalitya -.889 .294 9.137 1 < .01 .411 .231 .732 Sub. frequency -.346 .132 6.881 1 <. 01 .707 .546 .916 Sub. familiarity -.658 .190 12.073 1 < .01 .518 .357 .750 Length in phon. 1.322 2.133 .384 1 .536 3.750 .057 245.48 9 First-syl. freq. .040 .083 .236 1 .627 1.041 .885 1.224 MRC familiarity -.108 .103 1.084 1 .298 .898 .733 1.100 Zipf frequency -3.425 3.411 1.008 1 .315 .033 .000 26.048 Phonological ND .108 .112 .941 1 .332 1.115 .895 1.388 Constant 11.650 7.522 2.399 1 .121 114682.396 Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency;

Sub. familiarity = subjective familiarity, First-syl. freq. = first-syllable frequency, ND = neighborhood density aDefinition = 0

3.2.2.1. Cue modality

The binomial logistic regression revealed a significant decrease in the log odds of a phonological stage failure when the cue modality is a picture cue as compared to a definition cue (B = -.899,

SE =.294, Wald = 9.137, p < .01). More specifically, this indicates that the odds of having a retrieval failure at the phonological level is 59% less likely for picture cues than for definition cues.

3.2.2.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed a significant decrease in the log odds of having a phonological stage failure per unit increase in subjective frequency (B = -.346, SE =.132, Wald =

6.881, p < .01). Specifically, a phonological stage retrieval error is 29% less likely to occur when frequency increases. The regression also indicated a significant decrease in the log odds of having a phonological stage failure per unit increase in subjective familiarity (B = -.658, SE

=.190, Wald = 12.073, p < .01). This finding indicates that, for common nouns, a phonological stage failure is 48% less likely to occur when familiarity increases. Taken together, common nouns rated by participants with higher frequency and familiarity have an increased likelihood of successful phonological access.

3.2.2.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that length-in-phonemes (B = 1.322 SE =2.133, Wald =

.384, p = .536), first-syllable frequency (B = .040 , SE = .083, Wald = .236, p = .627), MRC familiarity (B = -.108 , SE = .103, Wald = 1.084 , p = .298 ), Zipf frequency (B = -.3.425 , SE =

3.411, Wald = 1.008 , p = .315), and phonological neighborhood density (B = .108 , SE = .112,

Wald = .941 , p = .332 ) were not significant predictors of phonological stage failure during common noun retrieval when accounting for the effect of the other predictors.

The overall results for models of phonological stage failure for common nouns are presented in Table 15.

3.2.3. Analysis of Proper Nouns

All models in this section only include proper noun trials. The results of the model presented in this section should be interpreted with caution, as the minimum sample size for detecting a small effect was not met (Button et al., 2013). A binary logistic regression with semantic stage failure regressed onto cue modality (picture vs. definition), subjective frequency, subjective familiarity, length-in-phonemes, and first-syllable frequency yielded an overall significant result of the model, χ2 (7, N = 1,344) = 150.604, p < .001, Nagelkerke’s R2 = .143. The results are summarized in Table 13.

Table 13

Regression results for cue modality, subjective frequency, subjective familiarity, length- in-phonemes and first-syllable frequency, as predictors of proper noun phonological stage failure

95% CI for OR B SE Wald df Sig. OR Lower Upper Cue modalitya -.151 .118 1.637 1 .201 .860 .682 1.084 SR frequency -.236 .054 19.121 1 < .001 .790 .711 .878 SR familiarity -.756 .085 79.231 1 < .001 .470 .398 .555 Length in phon. .611 .504 1.468 1 .226 1.842 .686 4.946 First-syl. freq. .009 .018 .232 1 .630 1.009 .973 1.046 Constant 1.810 1.560 1.346 1 .246 6.110 Note. CI = confidence interval, OR = odds ratio; Sub. frequency = subjective frequency;

Sub. familiarity = subjective familiarity; Length in phon. = length-in-phonemes; First- syl. freq. = first-syllable frequency, ND = neighborhood density aDefinition = 0

3.2.3.1. Cue modality

The binomial logistic regression showed that cue modality (B = -.151, SE = .118, Wald = 1.637, p = .201) was not a significant predictor of phonological stage failure during proper noun retrieval when accounting for the effect of the other predictors.

3.2.3.2. Semantically-related psycholinguistic variables

The binomial logistic regression showed a significant decrease in the log odds of having a phonological stage failure per unit increase in subjective frequency (B = -.236, SE =.054, Wald =

19.12, p < .001). Specifically, a phonological stage retrieval error is 21% less likely to occur when frequency increases. The regression also indicated a significant decrease in the log odds of having a phonological stage failure per unit increase in subjective familiarity (B = -.756, SE =

.085, Wald = 79.23, p < .01). This finding indicates that, for proper nouns, a phonological stage failure is 60% less likely to occur when familiarity increases. Taken together, proper nouns rated by participants with higher frequency or familiarity have an increased likelihood of successful phonological access.

3.2.3.3. Phonologically-related psycholinguistic variables

The binomial logistic regression revealed that neither length-in-phonemes (B = -.611, SE = .504,

Wald = 1.468, p = .226) nor first-syllable frequency (B = .009, SE = .018, Wald = .232, p =

.630), were significant predictors of phonological stage failure during proper noun retrieval when accounting for the effect of the other predictors.

The overall results for models of phonological stage failure for proper nouns alone are presented in Table 16.

Table 14

Summary of Direction of Independent Variables in Full Models

Semantic Stage Failure Phonological Stage Failure Noun typea N.S. ↓ Cue modalityb N.S. ↓ Noun type*Cue modality N.S. ↓ Subjective frequency ↓ ↓ Subjective familiarity ↓ ↓ Word-length-in-phonemes N.S. N.S. First-syllable frequency N.S. N.S.

Note. Downward arrows for continuous predictors (Subjective frequency and Subjective familiarity) indicate a significant decrease in the log odds of a semantic or phonological failure associated with increasing Subjective frequency and Subjective familiarity values. Downward arrows for categorical predictors (Noun type and Cue modality) indicate a significant decrease in the log odds of a phonological failure for common nouns relative to proper nouns and for picture cues relative to definition cues, respectively. N.S. = non-significant effect. aProper noun = 0, bDefinition = 0

Table 15

Summary of Direction of Independent Variables in Common Noun Models

Semantic Stage Failure Phonological Stage Failure Cue modalitya N.S. ↓ Subjective frequency N.S. ↓ Subjective familiarity ↓ ↓ Word-length-in-phonemes N.S. N.S. First-syllable frequency N.S. N.S. MRC familiarity N.S. N.S. Zipf frequency N.S. N.S. Phonological ND N.S. N.S.

Note. Downward arrows for continuous predictors (Subjective frequency and Subjective familiarity) indicate a significant decrease in the log odds of a semantic or phonological failure associated with increasing Subjective frequency and Subjective familiarity values. Downward arrows for the categorical predictor, Cue modality, indicates a significant decrease in the log odds of a phonological failure for picture cues relative to definition cues, respectively. N.S. = non-significant effect. aDefinition = 0

Table 16

Summary of Direction of Independent Variables in Proper Noun Models

Semantic Stage Failure Phonological Stage Failure Cue modalitya N.S. N.S. Subjective frequency ↓ ↓ Subjective familiarity ↓ ↓ Word-length-in-phonemes N.S. N.S. First-syllable frequency N.S. N.S.

3.3. Analysis of Demographic Factors

Semantic Stage Retrieval. The association between the demographic variables age and education and the percentage of semantic stage retrieval errors was not significant (r = .123, n = 52, p =

.477; r = .064, n = 52, p = .654). When the association between semantic stage retrieval errors was examined by noun type, age and education were not significant for common nouns (r =

.093, n = 52, p = .513; r = -.161, n = 52, p = .254) or proper nouns (r = .095, n = 52, p = .502; r

= .175, n = 52, p = .215). There was no statistically significant difference observed in the gender analyses for semantic stage retrieval errors for all noun types (t(50) = .148, p = .883), common nouns alone (t(50) = -.327, p = .745), or proper nouns alone (t(50) = .616, p = .541).

Phonological Stage Retrieval. The association between the demographic variables age and education and the percentage of phonological stage retrieval errors was not significant (r = .015,

n = 52, p = .917; r = -.105, n = 52, p = .460). When the association between phonological stage retrieval errors was examined by noun type, age and education were not significant for common nouns (r = -.129, n = 52, p = .362; r = -.212, n = 52, p = .131) or proper nouns (r = .153, n = 52, p = .279; r = .084, n = 52, p = .554). There was no statistically significant difference observed in the gender analyses for phonological stage retrieval errors for all noun types (t(50) = .644, p =

.522), common nouns alone (t(50) = .025, p = .981), or proper nouns alone (t(50) = 1.048, p =

.300).

4. Discussion

4.1. Overview

The purpose of this study was to evaluate the unique contribution of specific lexical features to word retrieval difficulties in older age for common and proper nouns at the semantic and phonological stages of retrieval. The TDH and prior empirical investigations suggest that common and proper nouns are differentially represented; however, the extent to which word retrieval difficulties, such as TOTs, occur as a function of psycholinguistic features known to interact with the semantic and phonological levels is not well understood. Few studies have systematically compared semantically- and/or phonologically-related features in a large set of common and proper nouns in the same experimental design. Additionally, no study to date has examined the effect of frequency and familiarity on proper noun retrieval, which is a critical next step in determining the lexical factors that influence semantic and phonological retrieval deficits in older adults. To address these research gaps, healthy older adults (n = 52) were tested on a

TOT-inducing stimulus set comprised of both noun types and controlled for cue modality

(presentation of picture or definition cue). The TOT paradigm included a fine-grained production analysis, referred to as the two-step approach that allows for a detailed characterization of word naming performance (Gollan & Brown, 2006).

The results of this study indicate areas of overlap and of distinction in the influence of psycholinguistic features on retrieval difficulties for both noun types at the semantic and phonological levels of processing. Interestingly, familiarity similarly modulated the likelihood of a retrieval failure at the semantic and phonological stages for both common and proper nouns, such that higher familiarity ratings were associated with successful retrieval. This finding is a novel contribution to the TOT literature as no study to date has examined the effect of familiarity

on the phonological-stage of lexical processing for both common nouns and proper names.

Additionally, higher frequency ratings led to greater success relative to lower frequency ratings at the semantic level for proper nouns, but not common nouns, which is consistent with the TDH and the notion of a distinct semantic architecture for common and proper nouns. Of further interest is the finding that cue modality, picture or definition cue, influences TOT likelihood in common nouns but not in proper names. This result uniquely implies that picture cues are less likely to lead to a phonological stage retrieval failure for common nouns than definition cues.

In the subsections to follow, the combined significance of the findings will be discussed relative to the extant literature. In addition, the theoretical and clinical significance will be addressed, followed by the limitations of the study. This section will conclude with suggestions for future research.

4.2. Influence of cue modality

Importantly, no study to date has examined how cue type influences the retrieval performance at the semantic stage of lexical processing. The findings of this study indicate that cue modality – the presentation of a picture or a definition cue – was not found to predict the likelihood of a semantic stage retrieval failure for either proper nouns or common nouns. According to the two- step approach, the abstract lexical representations of the target word are activated in the first stage of lexical retrieval. A failure at this stage signifies one of two potential issues: 1) the semantic representations corresponding to the target were never activated (“don’t know” response), or 2) the semantic representations did not activate the appropriate lexical representation (incorrect response) (Gollan & Brown, 2006; Hanley, 2011). The reader will recall that “don’t know” refers to instances when participants could not come up with the word but later recognized the word. The implication of this finding is that both pictures and definitions

interact with the semantic system to a similar extent, even though performance on pictures shows fewer TOTs than that on definitions. That is, this result suggests that the semantic information garnered from the verbal or visual stimuli does not differentially impact the likelihood of either a don’t know or incorrect response, termed a “Not GOT” in this present study, from occurring.

Current models of lexical retrieval are underspecified as they relate to the influence of semantic content obtained from verbal or visual means.

Interestingly, an effect of cue modality was observed at the phonological stage of retrieval such that picture cues reduced the odds of a TOT occurring. Additionally, this finding was significant in the full regression model that included both common and proper nouns as well as in the model that only included common noun targets. In order to further investigate this effect, an interaction term was added to the full regression model (cue type*noun type), and the likelihood of a phonological retrieval failure was reduced by 42% when common nouns were presented as pictures compared to trials when the targets were proper nouns presented via definition cues. One explanation, proposed by Brown and Nix (1996), for why picture cues are relatively less likely to lead to a TOT than definition cues is that the latter cue type may lead to an adverse effect of lexical competitors being activated from the semantic content presented in the definition. Persistent alternates are generally defined as non-target responses that repeatedly come to mind when attempting to retrieve the intended lexical item (e.g., Burke et al., 1991).

Although picture cues may also evoke competing alternates, they likely do so to a lesser extent than when the alternates are provided within a definition. For example, in a definition for the target word “reef,” the definition contains possible competitors, like “chain of rocks” and “coral”

(“What do you call a chain of rocks or coral at or near the surface of the water in an ocean;”

Burke et al., 1991, p. 574). Thus, a cue-related increase in competing alternates may result in more TOTs.

As mentioned in the Introduction (refer to Section 1), few studies have examined the effect of cue type on TOT occurrence (Brown & Nix, 1996; Hanley, 2011; Read & Bruce, 1982;

Salthouse & Mandell, 2013). Further, the effect of cue modality observed in the current study has not been found in prior TOT research. Of the aforementioned studies that have included a cue modality manipulation, only one study looked at common noun retrieval from picture and definition cues of the exact same target words (Brown & Nix, 1996). In their experiment, the authors found no effect of picture vs. definition cues on TOT rates for common noun retrieval in a group of younger and older adults. However, it should be stressed that their stimulus set was small (50 targets), and the results were based on the word naming performance of 20 older adults; therefore, the robustness of the effect is unclear. By contrast, the present investigation included a much larger stimulus set than that of Brown and Nix (1996).

The lack of an effect of cue modality for proper nouns supports the findings of prior TOT studies (Hanley et al., 2011; Read & Bruce, 1982; Salthouse & Mandell, 2013). One possibility for this result is that the special architecture of proper nouns creates an equivalence in cue modality. Recall that proper nouns are theorized to have multiple points of potential retrieval failure due to an increased number of one-to-one connections within both the semantic and phonological levels (Farrell, 2012). Thus, these bottlenecks will affect proper noun retrieval regardless of whether the cue type is a picture or definition. According to the TDH, common nouns, by contrast, are susceptible to transmission failures at stage two of lexical retrieval, which may explain why the semantic content from the verbal or visual modality has a differential influence on TOT propensity for common nouns relative to proper nouns. The semantic system

for common nouns includes more convergent connections between semantic nodes and the lemma node, so these connections can benefit from more semantic information provided by the cue. The findings of the present study support this hypothesis as common nouns presented as a picture cue reduced the odds of a TOT, but no such effect was observed for proper nouns.

Further systematic investigations will be necessary to disentangle the effect of cue modality on the retrieval of proper and common nouns at the semantic and phonological levels of lexical processing in older adults.

4.3. Influence of noun type

In this study, noun type (common vs proper) did not influence the likelihood of a semantic level retrieval failure (Not GOT); however, noun type did play a role in the likelihood of a phonological stage failure. That is, common nouns were 68% less likely than proper nouns to lead to retrieval difficulty at the phonological level. The lack of an effect of noun type at the semantic level is supported in the literature (Hanley, 2011; Juncos-Rabadán et al., 2010). Hanley

(2011) observed no difference between the two noun types on likelihood of a semantic stage failure when controlling for target word familiarity. In a separate study that did not control for familiarity, Juncos-Rabadán et al. (2010) also found similar rates of step one retrieval, termed

“success in semantic access” in their study, for both common and proper nouns in their youngest group of older adults. It is worth noting that a difference was observed in their oldest old group at the semantic level, wherein common nouns were more likely to result in retrieval success than proper nouns. The findings of the present investigation and those of Juncos-Rabadán and colleagues (2010) suggest that proper and common nouns exert a similar effect on semantic stage retrieval at least up through the ninth decade of life.

The finding that common nouns have a lower likelihood of resulting in a phonological stage failure when compared to proper nouns is well-established in the TOT literature (e.g.,

Burke et al., 1991; Evrard, 2002; Farrell, 2012; Fogler & James, 2007; James, 2006; Juncos-

Rabadán et al., 2010; Salthouse & Mandell, 2013, but see Hanley, 2011). In the present study, common nouns – when compared to proper nouns – were more than half as likely to not elicit a

TOT state. As mentioned in the Introduction (refer to section 1), this finding is expected in light of the TDH theory, since the TDH predicts that transmission failures are less likely for common nouns relative to proper nouns because the semantic structure of common nouns does not depend on the strength of one-to-one connections for lemma activation to occur. Therefore, the present investigation adds to the growing body of literature that common and proper noun retrieval differences are apparent at the phonological level of lexical processing, and to the much smaller literature on their lack of an effect at the semantic level.

4.4. Semantic-level psycholinguistic factors

In this section, the findings related to Specific Aim 1 are addressed. The focus is on the results related to the independent effect of frequency and familiarity on the semantic and phonological stages of lexical retrieval. As the reader will recall, the frequency and familiarity effects reported as subjective frequency and subjective familiarity refer to self-rated reports of the psycholinguistic features on a 5-point Likert scale.

4.4.1. Frequency effects at the semantic level

The present investigation found that in the main analysis, including both common and proper nouns, there was a significant, independent effect of subjective frequency such that an increase in the frequency rating (on a 5-point Likert scale) among participants was associated with a significant decrease in the odds of a semantic stage failure. A further examination of this

effect in separate analyses for common nouns and proper nouns indicated that there was no influence of frequency on semantic retrieval for common nouns, which was contrary to what was hypothesized.

Given a prior study that found an effect of frequency on semantic retrieval (Gollan &

Brown, 2006), it was originally predicted that as common noun word frequency decreased, there would be an increase in semantic stage failures. The lack of an effect in the current set of findings can be explained by predictions made by the TDH. More specifically, the convergent structure of the semantic system perhaps dampens the benefit that word use might impart to lemma retrieval. That is, lemma activation for common noun retrieval is not dependent on one- to-one connections, which is the type of architecture that shows a frequency-related advantage in the likelihood of retrieval success. However, in Experiment 1, Gollan and Brown (2006) found that “easy” common nouns, which tended to be higher frequency on average and higher familiarity than “difficult” common noun targets, resulted in significantly fewer semantic-level failures than difficult words. One reason the researchers might have found a frequency effect for semantic retrieval is that all their words tended to be low frequency, with few high frequency targets included in the stimulus set. Therefore, it could be that within low frequency there are words that are more or less likely to lead to a semantic retrieval failure. As well, the easy and difficult targets differed on self-rated familiarity, which could have also played a role in the likelihood of a semantic-level difficulty. Given the differences in the stimulus sets of the present study and Gollan and Brown (2006) experiment, it would seem that the results are complimentary rather than in conflict with one another.

With respect to proper noun retrieval, subjective frequency had a significant effect on retrieval at the semantic level. These findings are significant even when accounting for the

possible contribution of familiarity, word length in phonemes, and first-syllable frequency. This is a novel contribution to the TOT literature, as no TOT study to date has examined the independent effect of frequency on proper noun retrieval at the semantic level. As mentioned in the introduction, no study has examined the effect of frequency on TOT rates during proper noun retrieval due to the fact that objective frequency counts for proper nouns are unavailable in

American English (Brédart, 2017). Taken together, an effect of frequency at the semantic level is observed in the retrieval of proper nouns and not common nouns due to the unique architecture of the semantic system of proper nouns, consistent with the TDH.

4.4.2. Frequency effects at the phonological level

As hypothesized, higher subjective frequency ratings lessened the likelihood of common noun and proper nouns retrieval failure at the phonological level (i.e., TOT occurrence). The effect of frequency on common nouns’ TOT propensity is relatively well-established in the literature (e.g., Astell & Harley, 1996; Burke et al., 1991; Gianico-Relyea & Altarriba, 2012;

Gollan & Brown, 2006; Gollan & Silverberg, 2001; Gonzalez, 1996; Harley & Bown, 1998;

Vitevitch & Sommers, 2003). These results are in line with the TDH, which suggests that there is a direct relation between how often a word is accessed and how much activation is being sent from the lemma level to the corresponding phonological nodes. Words that are not frequently used will have reduced opportunities to send activation between these one-to-one connections, which are theorized to weaken with age (Abrams & Davis, 2016).

An effect of subjective frequency at the phonological level retrieval for common nouns builds upon the existing psycholinguistic literature. Recall that, Newman and German (2005) argued that frequency, as a psycholinguistic factor, may have more of an effect at the semantic level of lexical processing rather than the phonological level, since frequency is linked to the

usage of a word. Indeed several other studies analyzing TOT states (Gollan & Brown, 2006) or speech errors in healthy individuals (Harley & MacAndrew, 2001; Vitkovitch & Humphreys,

1991) have found that frequency additionally affects the semantic stage of retrieval. However, the current findings, as they relate to common noun retrieval, are consistent with a seminal study conducted by Jescheniak and Levelt (1994), which posited that frequency exclusively influences the phonological stage of retrieval and not the semantic stage (see Kittredge et al., 2008 for a review).

The significant effect of subjective frequency on proper noun retrieval is a unique contribution of the present investigation to TOT research, as it is the only study that has investigated this psycholinguistic phenomenon for English proper nouns. The findings are consistent with those of Bonin and colleagues (2008), who found that objective frequency ratings of the names of French celebrities correlated negatively with TOT rate among young adults.

Thus, the current study extends this finding to a group of healthy older adults. In sum, high- frequency words are less susceptible to TOT states than low-frequency targets for both common and proper nouns.

4.4.3. Familiarity effects at the semantic level

Prior research has called for further investigation into the role of other psycholinguistic factors that might, independent of frequency, modulate TOT occurrence (e.g., Farrell & Abrams,

2011; Gollan & Brown, 2006; Hanley, 2011). Thus, an objective of the present study was to determine the potential influence of target familiarity on semantic and phonological processing.

Word frequency and familiarity are often correlated with one another (Tanaka-Iishi & Terada,

2011), and these psycholinguistic factors are moderately correlated in the current study (r = .546, n = 52, p < .001). However, the two features are dissociable from one another. Specifically, high

frequency words are generally also high familiarity; however, high familiarity words are not necessarily high frequency. Some examples of words in the current dataset rated as high familiarity and high frequency (rated at the highest value of 5 on both scale) include Donald

Trump and John Lennon, as well common nouns like mushroom and thermometer. Targets that were rated as high familiarity (given a value of 5) and low frequency (given a value of 2) include proper nouns like Alfred Hitchcock and Ella Fitzgerald and common nouns such as banjo and transom. As hypothesized, a higher subjective familiarity rating was associated with a significant reduction in the odds of a retrieval failure at the semantic stage for both common and proper nouns. This effect is consistent with prior studies examining the effect of target familiarity for common nouns (Gollan & Brown, 2006; Harley & Bown, 1998) and proper nouns

(Hanley & Chapman, 2008). Importantly, this is the first investigation to analyze the unique contribution of familiarity to the semantic stage of lexical retrieval for both common and proper nouns within the same experimental design. Further, these findings are the first to indicate that older adults are sensitive to an effect of familiarity at the semantic level during proper noun naming.

4.4.4. Familiarity effects at the phonological level

As predicted, an increase in subjective familiarity values reduced the likelihood of a phonological stage retrieval failure for both common and proper nouns. Diary studies (Burke et al., 1991; Cohen & Faulkner, 1986; Ecke, 2004) and laboratory investigations (Gollan & Brown,

2006; Hanley & Chapman, 2008) have observed an effect of familiarity on TOT occurrence; however, the results of the two studies’ designs indicate an opposing direction of the effect, with highly familiar words more likely to induce TOTs in the former and less likely in the latter. The finding of the present study that an increase in familiarity reduced the likelihood of a TOT is

consistent with the literature taking a laboratory-based approach to TOT research. One potential explanation for the discrepant findings between the two paradigms is that diary studies likely reflect the use of well-known and relatively familiar vocabulary. Laboratory investigations, by contrast, do not necessarily account for a participant’s level of familiarity with the stimulus set.

Given a two-stage interpretation of word production and the predictions of the TDH, familiarity, as a psycholinguistic construct, bolsters one-to-one connections vulnerable to age-related weakening in the semantic system as well as the phonological system for both common and proper nouns. The analyses in the present investigation are the first to clarify the independent contribution of frequency and familiarity to common and proper noun retrieval in aging at the semantic and phonological level. Further, these findings confirm the validity of taking a two-step approach to word retrieval analysis. To summarize, higher familiarity ratings reduced the likelihood of semantic and phonological failures for both common and proper nouns.

4.5. Phonological-level psycholinguistic factors

This section of the discussion addresses Specific Aim 2 and focuses on the findings related to the independent effects of word length, first-syllable frequency, and neighborhood density on the semantic and phonological stages of lexical retrieval for both common and proper nouns.

4.5.1. Phonological psycholinguistic factors at the semantic level

One of the objectives of the present investigation is to determine whether retrieval failures at the semantic level are modulated by lexical characteristics hypothesized to affect the phonological stage of word retrieval: word length (i.e., total number of phonemes in the target), first-syllable frequency, and neighborhood density. As hypothesized, there was no effect of any of the phonological psycholinguistic factors on the semantic stage of lexical retrieval, regardless of noun type. The scarcity of studies on this specific question, however, might be attributed to

the fact that there is no theoretical justification for assuming an effect of word length (in phonemes) on semantic retrieval for common nouns or proper nouns, as word-form activation occurs at the phonological stage of retrieval. In their analysis of proper noun retrieval failures at the semantic level in young adults, Hanley and Chapman (2008) came to a similar outcome regarding the effect of word length.

Similarly, there is little theoretical support for an effect of first-syllable frequency at the semantic level for either noun type. The influence of first-syllable frequency is largely attributed to the phonological stage of processing. Thus, the present findings are consistent with this prevailing idea.

Another aspect of TOTs that has been investigated for its influence on TOT incidence is neighborhood density. The reader may recall that the analysis included only the common noun trials because objective neighborhood density values do not exist for proper nouns. Following the same line of reasoning put forth for word length and first-syllable frequency, there is no theoretical justification for the relevance of neighborhood density on the semantic stage of retrieval, and this is in fact what I found. Indeed, the present study offers empirical evidence against the involvement of word length, first-syllable frequency, and neighborhood density on the likelihood of a retrieval error (Not GOT) at the semantic level.

4.5.2. Phonological psycholinguistic factors at the phonological level

Unexpectedly, the present study found no effect of word length or first-syllable frequency on TOT occurrence for common and proper nouns. The TDH predicts that an increase in length- in-phonemes would correspond to an increase in opportunities for reduced transmission of activation due to the additional phonological nodes requiring activation. This dissertation’s finding that word-length-in-phonemes had no effect on retrieval at the phonological level for

common nouns is thus inconsistent with those of Hanly and Vandenberg (2010), who observed a higher TOT rate in children with and without dyslexia for common nouns containing more phonemes than those containing fewer phonemes. However, recall that Hanly and Vandenberg

(2010) did report a significant interaction such that the word length effect was more pronounced for low frequency targets than high frequency ones. In the present study, other psycholinguistic variables, such as frequency, were controlled for in the regression analyses in order to isolate the effect of word length on TOT likelihood. The current findings suggest that when controlling for the known inverse association between word length-in-phonemes and word frequency, which was small but significant in the present study (r(4,375) = -.17, p < .001), the former does not modulate TOT propensity during common noun retrieval (e.g., Hodgson & Ellis, 1998;

Mortensen et al., 2006; Zipf, 1935).

Although word length is among the psycholinguistic factors thought to influence ease of word retrieval (Meyer, Roelofs, & Levelt, 2003; Luce & Pisoni, 1998), there is relatively little research on its effect on TOT production in aging. Only one such study to date examined the role of proper noun word length on TOT rate, albeit in young adults (Hanley & Chapman, 2008). The authors noted that three-name proper nouns were more likely to result in a phonological-level deficit than two-word proper nouns. The researchers controlled for familiarity of the targets but not for frequency (given the previously discussed lack of objective frequency values). Given the role of frequency on TOT rate observed in the present study, it is possible that Hanley &

Chapman’s (2008) findings are confounded with word length. Indeed, prior studies on object naming latencies found that when frequency and familiarity are controlled for, there is no effect of word length on response times (see Meyer et al., 2003 for a review). Therefore, future research is needed to assess whether word length effects are independent contributors to

phonological retrieval difficulties, or whether, as the present study suggests, frequency and familiarity are predictors of TOT likelihood in healthy older adults.

An effect of first-syllable frequency on phonological retrieval failure was also predicted but not upheld by the results for either common or proper nouns. In the general literature, there are far fewer studies on this psycholinguistic feature compared to the other features included in the analyses. Prior studies controlled for pre-experimental familiarity of both noun types and found that high frequency first syllables reduced TOTs for common nouns and increased TOTs for proper nouns in older adults (Farrell & Abrams, 2011; Farrell, 2012). According to the present findings, it does not appear that syllable frequency predicts the likelihood of phonological failure for proper nouns or commons nouns over and above the unique contributions of frequency and familiarity.

As hypothesized, neighborhood density did not influence the likelihood of a phonological stage failure for common nouns. Prior research indicates that young adults experience a higher

TOT incidence for words from sparse neighborhoods than from dense neighborhoods (Harley &

Bown, 1998; Vitevitch & Sommers, 2003). However, in those studies a main effect of neighborhood density was not observed among older adults; rather, TOT propensity increased for words from a sparse neighborhood that also had a low neighborhood frequency. The present study did not account for neighborhood frequency values for the common noun targets, which could account for differences between this study and Vitevitch & Sommers (2003). One interpretation of the current result within the TDH framework is that both sparse and dense neighborhoods experience age-related weakening in the transmission of priming that reduces the subsequent feedback between the phonological and lemma levels. The beneficial feedback is greater for common nouns from denser neighborhoods than words from sparser neighborhoods.

Thus, age-related weakening at the phonological level will reduce this beneficial feedback and result in a similar effect of neighborhood density on TOT propensity (at least for common nouns) for sparse and dense neighborhoods.

4.6. Aging effects

The present findings suggest that there is no effect of age on the likelihood of a semantic or phonological stage retrieval error for all noun types or when noun types were analyzed separately. This is not altogether unsurprising. Certainly, age effects at the phonological stage of retrieval have frequently been observed in the TOT literature for both common nouns and proper nouns, such that older adults experience more TOTs than younger adults (e.g., Abrams, Trunk, &

Margolin, 2007; Brown & Nix, 1996; Burke et al., 1991; Dahlgren, 1998; Farrell, 2012; Heine,

Ober, & Shenaut, 1999; James & Burke, 2000; Ouyang et al., 2020; Rastle & Burke, 1996;

Vitevitch & Sommers, 2003; White & Abrams, 2002) and middle-aged adults (M = 30-49 years;

Dahlgren, 1998). However, the majority of these studies only compare older adults to a younger participant group, usually in their 20’s, while the present study examined performance among older adults (54-89 years, mean age = 68, SD = 9.2).

Only two studies have been conducted to date that have examined performance among older adults relative to an old-old group (Heine et al., 1999; White & Abrams, 2002). Heine and colleagues (1999) examined two groups of older adults: the first between 60 and 74, and the second between 80 and 92 and found a significant effect of age on the occurrence of TOTs, with the old-old individuals committing more retrieval errors at the phonological level only than the old group. By contrast, White and Abrams (2002) found no such effect of age between their groups of older adults (60-72 years vs. 73-83 years). The difference between the findings in the present study as well as those of Heine et al. (1999) and White and Abrams (2002) may be

attributed to distinct approaches to calculating TOT errors. The lack of an age effect in the present investigation is in line with the findings of White and Abrams (2002). The average age of participants in the present investigation was 68, and the age range of the present study crosses the two age demographics of White and Abrams (2002) study as compared to Heine et al. (1999).

Specifically, only four participants in the current study fell into the old-old age group of Heine et al. (1999), while 15 participants fit into the oldest age group of White and Abrams (2002).

Therefore, an age effect within healthy older adults may not occur at the phonological level until one’s 80s. For the participants in the present study, there is not an age-related difference in phonological level retrieval difficulty among a group of healthy older adults who are, on average, over a decade younger than 80.

At the semantic level, research has commonly examined age effects between young adults and older adults, but minimally within older adults. The current study is the first to look at an effect of age on semantic retrieval within a group of healthy older adults using the calculation proposed by Gollan and Brown (2006). Neither common nouns nor proper nouns were associated with an age-related increase in semantic-level naming difficulty. One prior study found that, for common noun retrieval, there was no difference in “don’t know” responses between their two groups of older adults: 19% for young-old and 18% for old-old (White & Abrams, 2002). “Don’t know” responses, here, are being interpreted as a proxy for Not GOTs, which is indicative of a semantic stage failure, since I carefully excluded all ‘never known’ items, which White and

Abrams (2002) did not. Additionally, between younger and older adults, a reverse age effect has been found wherein older adults are less likely than younger adults to commit a semantic error for both common and proper nouns (e.g., Gollan and Brown, 2006; Juncos-Rabadán et al., 2010).

However, other research has found no effect of age (mean age in years: young adult = 20.63;

older adult = 65.78) at the semantic level for proper noun retrieval (Ouyang et al., 2020). These combined results are consistent with an age-related retrieval advantage for semantic retrieval, since older adults have extensive word knowledge attributable to their years of experience, and thus have fewer opportunities for retrieval failure (Gollan & Brown, 2006; White & Abrams,

2002). One hypothesis as to why age-effects are less likely during semantic retrieval is that word knowledge or experience can be likened to a frequency effect, which may be more likely to affect the first stage of retrieval (semantic) than the second stage of retrieval (phonological). That is, knowledge of words may enrich semantic representations (Gollan & Brown, 2006). Based on the findings of the present study, this concept holds true for both common nouns and proper nouns in a group of healthy older adults.

4.7. Clinical Implications

The findings of this study have direct clinical implications for assessment and treatment of age- related lexical retrieval changes in healthy individuals as well as in neurological disorders, such as those with dementia, primary progressive aphasia, and aphasia, whose naming difficulties are one of the most functionally debilitative consequences of their condition. This study contributes novel findings to the literature in that cue modality appears to modulate TOT rates for common nouns. With this in mind, both clinicians and researchers alike should select the modality of presentation of common noun stimuli with awareness of the differential influence of picture and definition cues on word naming performance. Careful consideration of cue type is necessary in the neuropsychological assessment of naming abilities, as common noun naming tasks that include picture cues may not be sensitive enough to show impairments in comparison to naming tasks cued from definitions. Attention to the cue type is all the more important as most assessments primarily, if not exclusively, employ picture cues. By contrast, TOT researchers

might choose to utilize definition-cued common nouns rather than picture-cued common nouns to increase opportunities for retrieval failure. In addition, the Complexity Account of Treatment

Efficacy (CATE; Thompson, 2003), suggests that training complex or challenging targets may generalize to less-complex targets that share a similar set of processes. The results of the present study suggest that definition cues are more challenging to retrieve from than picture-cues (i.e., definitions result in more TOTs than pictures). Therefore, this account supports the training of definition cues in lexical retrieval treatment to promote improved word naming ability in older adults. A TOT-inducing stimulus set would allow investigators to examine additional causes of retrieval breakdowns for common and proper nouns, such as target word characteristics and may offer clinicians relevant information for the initial use of intervention strategies.

Another goal of this dissertation was to identify the psycholinguistic features of words that are more or less likely to lead to retrieval failures at the semantic and phonological levels.

The finding that successful phonological access is more likely for high frequency and high familiarity targets for both common and proper nouns can assist clinicians in selecting therapeutic materials used in the treatment of word-naming deficits. Although word characteristics such as frequency and neighborhood density are commonly considered when selecting treatment words (Storkel, 2018), degree of familiarity with the target is rarely taken into account. The results of this study point to the importance of incorporating psycholinguistic characteristics into stimulus selection in order to implement evidence-based practice.

4.8. Limitations and Future Research

This research offers novel findings on the independent effects of frequency and familiarity on the likelihood of retrieval success and failure at the semantic and phonological stages of lexical processing. Nevertheless, due to limitations in the current design, additional research will be

necessary to bolster the findings regarding the influence of psycholinguistic features on semantic- and phonological-level retrieval. Ideally, one would try to equate the picture and definition cues for semantic content. The present investigation did not establish this for the two cue types. This drawback is more likely to have an effect on proper names than common nouns because proper name targets for pictures only include the individual’s face. By contrast, the definition included biographical content that was not controlled for in this study. Further, the semantic content within proper and common noun definitions may have differed such that the former is semantically richer than the latter. As mentioned earlier in section 4.3, definition cues are more likely than pictures to result in TOTs, and proper name definitions contained more semantic content than common noun definitions. Thus, it cannot be ruled out that the difference between the two noun types at the phonological level is attributable, at least in part, to the nature of the semantic content contained in the cue.

It is important to note that all findings related to proper noun retrieval refer exclusively to proper names (Bette Midler) and not to places (Philadelphia). It is possible that famous place names might be more or less vulnerable to retrieval failures at the semantic and phonological levels compared to famous people. Prior research suggests that proper names and places may be differentially represented in the brain, with names for people relegated to the left anterior temporal lobe and place names to the right anterior temporal lobe (Ross & Olson, 2011).

Therefore, it would be of particular use for future studies to expand the present stimulus set to evaluate potential differences in the retrieval of both types of proper nouns in older adults.

The distribution of common and proper nouns across the sublists varied – with some sublists containing as few as 8 common nouns, for example, and others containing upwards of

24. This is a methodological drawback that occurred when the reduced processing load of

Google Forms became unmanageable and necessitated the creation of additional sublists.

Although it would have been ideal to have a balanced distribution of common and proper nouns in each sublist, the analyses in this dissertation were run on trial-level comparisons and not participant-level comparisons. Therefore, the findings are unlikely to be the result of participant- level differences in proper noun and common noun assignment.

With regard to first-syllable frequency, the measure was calculated from the CELEX database and reflects the combined frequency of all common nouns with that initial syllable. It is important to emphasize that proper names rarely occur within the database. Thus, proper name initial syllable values were obtained from common noun syllable frequencies. This calculation procedure was consistent with prior studies (Farrell, 2012; Farrell & Abrams, 2014). It is not known whether first-syllable frequency values among common nouns are the same as first- syllable frequency values among proper names. Therefore, it may be worthwhile for researchers to examine this question, either by employing a subset of proper names within the CELEX database or through other means. The implication for the current study is that the first-syllable frequency values for proper names may not truly reflect the syllable’s use among proper names.

Another recommendation for future research is to alter the wording utilized in the frequency and familiarity self-rating scales for common and proper nouns. The self-rating scales were modified from Farrell (2012) and are potentially problematic as they conflate frequency and familiarity in the written descriptions ascribed to each point on the Likert scale. For example, on the 5-point familiarity rating scale, 5, which corresponds to “very familiar,” is described as “I hear and use the word regularly and could provide a clear definition of the word.”

The use-related wording muddies the clarity of the familiarity content (i.e., that the participant knows the target well enough to be able to define it). The conflation of frequency and familiarity

in the rating scales did not appear to negatively influence the detection of frequency and familiarity effects, as both psycholinguistic factors contributed uniquely to explanation of the dependent variables in a number of models.

This dissertation exclusively referred to the TDH as the theoretical framework from which to interpret outcomes, as it is the most developed model for explaining age-related TOT increases. Furthermore, the TDH offers an account of why TOT incidence is greater for proper noun retrieval relative to common noun retrieval. However, alternative explanations of TOT occurrence in aging have been proposed, such as the Inhibition Deficit Hypothesis (IDH; Zacks

& Hasher, 1994). This hypothesis emphasizes the role of inhibitory processes in lexical retrieval

– specifically in reducing competition from alternate words when selecting the intended target

(Abrams & Davis, 2016). According to the IDH, TOTs occur because of blocked retrieval from competing alternates. Assuming that inhibition as a cognitive construct declines with advancing age, older adults experience an increase in TOTs due to reduced inhibitory skill relative to younger adults. Research has found little support for the IDH, as TOTs occur in older adults without the presence of persistent alternates. Even when inhibition performance is accounted for,

Ouyang and colleagues (2020) found an asymmetric effect of age on TOT rates in their group of younger and older adults, such that the latter age group experienced more TOTs than the former age group. As well, Higby et al. (2019) did not observe an effect of inhibition on naming accuracy or response times in their group of older adults. Future studies should continue to explore alternative accounts of TOT etiology in aging.

As mentioned in the statistical analysis of common nouns (refer to section 3.2.2), the minimum sample size for detecting a small effect was not met. Therefore, the models presented

in that section should be carefully interpreted. Future studies should investigate the veracity of the effects in those models.

4.9. Conclusion

In sum, this study examined the differential influence of psycholinguistic features on the semantic and phonological levels of common nouns and proper names in a group of healthy older adults in order to understand the locus of retrieval failures for both noun types. The findings provide additional evidence in support of the TDH. The distinct semantic and phonological architecture of common nouns and proper names, as outlined in the TDH, directly relate to the likelihood of successfully accessing semantic and phonological representations for both noun types. Further, a novel contribution of this dissertation is the role that psycholinguistic factors such as frequency and familiarity have in modulating the likelihood of retrieval success and failure at both stages of lexical processing. The pattern of results indicates that a frequency- related advantage in semantic and phonological processing favors one-to-one connections, which accounts for the observed overall benefit at both stages for proper names and at the phonological stage alone for common nouns. This research also pointed to a cue-related retrieval advantage at the semantic level for common nouns retrieved from picture cues that speaks to the architecture of common nouns: convergent connections are more efficient at extracting semantic content from visual content that tends to be less semantically-rich than verbal content. Current models of lexical retrieval, including the TDH, are underspecified as they relate to the interaction of semantic content obtained from pictures and definitions at both stages of lexical processing.

Overall, these findings provide evidence that retrieval difficulties for common nouns and proper names are influenced by psycholinguistic properties, such as frequency and familiarity, that differentially interact with the unique semantic and phonological architecture of both noun

types. This research contributes to a growing body of literature that aims to characterize the overall representation of common and proper names in the semantic and phonological systems of the mental lexicon.

Appendices

Appendix A. All stimuli alongside their associated psycholinguistic properties

Freq. 160 1242 172 1242 87 87 87 87 87 87 87 87 First Syl. First 3163 3163 87 269

Phono. ND 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2

0 0 0 3.05 0 Frequency 3.14 3.12 2.87 0 0 0 0 0 0 0 0

Familiarity 394 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

11 11 Phon. Length 10 8 10 7 8 8 5 12 8 8 9 13 11 12

5 5 5 3 3 4 4 4 Syl. Length 4 4 4 4 4 6 4 2

Letter Letter Length 11 9 11 11 14 8 8 9 10 13 16 14 10 15 6 8

Word Count 1 1 2 2 2 2 2 1 1 2 2 2 2 2 2 2

Alan Bates Alan Al Capone Alda Alan Al Pacino accelerator Brody Adrien Christie Agatha acupuncture accordion Word Alan Greenspan Alan Einstein Albert MacGraw Ali Alan Davies Alan Alanis Morissette Alanis Hitchcock Alfred almond

3163 276 139 276 276 276 First Syl. First Freq. 126 126 3163 126 139 276 276 276 276 1242 276

Phono. 0 0 0 3 0 0 0 ND 3 1 0 0 2 0 0 0 0 0

Frequency 3.97 4.35 0 1.89 3.87 0 0 0 0 1.59 0 0 0 0 0 3.9 0

0 0 0 0 543 Familiarity 0 0 0 0 0 0 499 0 458 0 0 0

Length 12 11 10 9 8 Phon. 4 9 8 4 12 10 9 12 14 10 7 5

4 2 3 6 4 5 5 5 Length 5 3 5 4 Syl. 2 3 2 5 2

5 9 13 9 6 14 17 12 9 10 10 10 14 5 13 12 10 Letter Letter Length

2 3 2 2 2 2 1 2 Word Count 1 1 2 1 1 0 2 2 1

Andie Macdowell Andie Andy Griffith Andrew Lloyd Webber Andrew Lloyd anchor amber Earhart Amelia ampersand ambulance Word Andy Warhol Baker Anita Huston Anjelica Anna Paquin Anne Bancroft Anne Frank anemometer Jolie Angelina ankle

1242 8 8 269 269 3163 First Syl. First Freq. 276 276 276 276 276 8 276 172 3163 269 269

0 Phono. 0 0 0 0 1 8 0 ND 0 0 0 0 10 0 0 0 0

3.43 0 3.86 0 3.89 0 0 3 0 2.75 Frequency 0 0 0 0 0 2.5 2.81

534 490 0 0 0 0 Familiarity 0 0 0 0 516 0 0 0 302 376 0

9 12 10 5 5 Phon. Length 8 13 5 13 5 17 3 8 8 7 9 12

4 4 3 2 5 2 3 6 4 5 5 4 2 2 Syl. Length 4 2 4

5 9 12 5 14 5 5 20 Letter Letter Length 12 11 14 14 12 6 12 9 9

Word Count 2 2 2 2 2 1 1 1 1 1 2 2 2 2 1 1 2

apron Word Franklin Aretha arrow armor Arnold Schwarzenegger antler Anthony Hopkins Anthony Quinn anvil Anthony Perkins Anne Hathaway Lennox Annie Arthur Miller artichoke Art Carney Art Garfunkel asparagus

419 126 1 151 13 11 First Syl. First Freq. 1242 9 1242 1242 486 486 486 1242 38 486 5

0 Phono. 0 0 3 0 0 0 5 ND 1 0 0 26 4 10 0 0 0

3.53 1.77 2.69 3.69 0 0 0 2.5 4.07 2.59 Frequency 3.6 3.02 1.99 3.89 4.4 0 0

397 0 0 0 0 0 Familiarity 0 0 511 0 399 257 0 0 0 0 461

5 4 6 10 7 Phon. Length 8 7 12 3 4 3 9 7 5 9 6 6

2 2 3 4 1 2 3 1 2 4 2 3 4 3 Syl. Length 3 6 2

11 6 8 9 3 15 6 3 Letter Letter Length 9 5 8 5 8 13 5 9 7

Word Count 1 1 1 1 1 2 2 1 1 2 3 1 1 2 1 1 1

avalanche Word awl Ayrton Senna awning axe Audrey Hepburn atomizer audience Augusto Pinochet attic astronaut atlas bacon badminton B. King Ruth Babe bagpipe

416 27 11 416 151 First Syl. First Freq. 3 93 416 93 17 773 200 200 17 200 416 416

0 Phono. 0 0 3 0 0 7 11 0 0 ND 2 0 14 0 0 46 1

3.74 3.59 2.81 4.03 0 0 2.41 3.2 4.31 3.46 Frequency 3.86 2.41 3.12 3.22 3.08 0 0

293 424 0 514 0 0 0 0 0 0 0 0 487 Familiarity 0 0 277 0

7 7 5 8 10 5 5 6 Phon. Length 7 10 4 8 10 7 3 5 5

2 1 2 3 1 3 2 5 3 3 2 2 5 5 2 3 Syl. Length 3

12 10 6 6 14 6 5 8 7 7 Letter Letter Length 7 7 8 5 11 3 5

1 Word Count 1 1 1 1 1 2 2 1 1 1 1 2 1 1 2 1

Word barrel Manilow Barry barge barnacle baptism Hershey Barbara barber banister Obama Barack banjo bandana balcony baster bat baton bayonet Barry White Barry

11 126 43 43 43 First Syl. First Freq. 773 126 6 126 56 314 25 25 25 161 11 11

0 Phono. 0 12 8 0 0 0 0 0 0 ND 25 7 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 2.96 2.65 4.1 3.68 2.44 2.92 0

0 0 0 0 0 0 0 470 211 0 0 0 0 Familiarity 538 0 0 0

6 4 4 6 8 9 9 7 Phon. Length 4 5 14 10 8 10 10 8 7

3 3 3 4 7 4 4 4 3 2 2 2 2 3 3 2 Syl. Length 1

11 11 7 10 10 11 15 13 6 9 Letter Letter Length 7 6 5 6 7 11 9

1 Word Count 1 1 1 1 1 1 2 2 2 2 2 2 1 1 2 1

Affleck Word Davis Bette Midler Bette Benito Mussolini Benito Sanders Bernie belfry Ben Kingsley Ben beard bellows beaver beaker bazooka Beyonce Clinton Bill Cosby Bill Gates Bill Betty Grable Betty

42 105 8 17 58 First Syl. First Freq. 43 43 43 223 24 105 42 42 42 42 42 42

0 Phono. 0 0 0 2 0 0 0 0 14 ND 0 0 0 0 0 15 0

0 0 0 0 0 0 2.86 0 3.84 3.31 Frequency 0 0 0 3.2 3.07 0 0

0 0 0 466 0 0 0 0 0 0 0 0 0 Familiarity 0 0 0 0

6 10 10 6 6 8 10 14 Phon. Length 7 9 8 9 9 10 4 4 4

2 1 2 5 3 3 4 4 3 4 4 2 3 2 3 6 Syl. Length 3

10 12 4 12 7 8 9 10 9 17 Letter Letter Length 10 14 9 10 9 4 4

3 Word Count 1 3 2 1 1 2 2 2 2 2 2 2 1 1 2 1

Washington

Word Gentry Bobbie Bobby Darin Bob Marley Bob Newhart bleachers Bob Dole Bob Dylan Billy Joel Billy Bo Hopkins binoculars Billie Jean King Jean Billie Bill Murray Bill bolo bolt Bono T. Booker Bobby Kennedy

100

0 0 0 35 41 First Syl. First Freq. 113 14 3 3 1 101 3 12 98 0 3 0

0 Phono. 0 0 0 17 0 9 15 0 0 ND 0 21 0 0 0 0 4

4.66 0 0 3.68 0 0 0 0 0 4.15 Frequency 2.97 3.89 0 0 2.75 3.78 0

0 579 0 0 0 0 547 0 0 551 0 561 547 Familiarity 0 0 0 0

7 7 7 4 12 4 4 10 Phon. Length 9 4 11 9 7 6 9 4 13

3 3 1 2 4 3 1 2 3 2 2 1 1 4 1 4 Syl. Length 3

9 8 16 5 13 6 14 11 7 10 Letter Letter Length 9 8 9 8 5 11 5

2 Word Count 1 1 2 2 1 1 2 1 2 2 1 2 2 2 2 1

Word broom Dern Bruce Brigitte Bardot Brigitte Peters Brock braille Vaccaro Brenda bridge Brad Davis Brad brake Brad Pitt Brad bracelet boomerang Bruce Springsteen Bruce Willis Bruce brush Adams Bryan Bruce Lee Bruce

101

795 795 412 795 73 795 First Syl. Freq. 407 54 0 0 0 17 407 407 5 9 71

0 Phono. ND 0 0 1 0 0 2 0 8 21 0 0 0 9 17 2 0

4.45 3.47 4.31 3.12 3.85 1.59 3.27 0 3.7 0 Frequency 0 3.08 0 0 0 4 3.74

341 421 0 0 0 0 Familiarity 0 425 573 400 0 0 0 0 481 487 0

Phon. Length 11 6 5 5 6 7 5 8 9 3 10 7 3 10 6 10 15

3 5 2 3 2 1 4 Syl. Syl. Length 4 2 2 2 3 4 2 3 3 5

8 7 16 6 6 9 4 10 Length 10 Letter Letter 12 4 13 12 9 6 5 11

1 1 Word Count 2 1 2 2 2 1 1 1 1 1 1 2 1 2 1

cactus calendar Word button calculator cage Burt Burt Young Burt Burt Lancaster butler butterfly Burt Burt Reynolds Buddy Hackett buoy camouflage Calista Flockhart camel caliper Cameron Cameron Diaz

102

38 795 249 795 795 795 First Syl. First Freq. 795 767 154 249 249 249 795 795 38 249 249

0 Phono. 7 2 0 0 0 0 0 ND 1 0 0 0 0 1 0 12 0

3.16 0 2.87 2.07 0 1.99 4.33 2.8 2.07 1.89 Frequency 3.94 3.55 3.24 2.77 0 0 0

0 0 0 0 415 0 Familiarity 441 0 0 0 0 0 498 490 0 396 0

4 8 8 13 7 Phon. Length 5 6 8 8 8 10 9 9 7 9 8 5

2 3 4 3 4 3 3 4 2 2 3 3 5 3 Syl. Length 2 3 3

9 9 6 8 12 10 9 12 Letter Letter Length 6 5 7 8 9 13 8 9 8

Word Count 1 1 1 1 2 2 2 1 1 2 1 1 1 1 1 1 2

carousel Word Fisher Carrie Grant Cary cartilage cartographer Carlos Santana Carlos canteen Lewis Carl Carole King Carole cardigan cannon canoe catacomb catamaran castanets castle catapult

103

3 103 62 62 62 62 First Syl. First Freq. 0 795 795 16 316 497 214 103 15 767 11

0 Phono. 0 0 0 0 0 23 0 ND 0 0 1 18 1 1 2 0 0

2.81 3.56 2.87 3.16 3.41 3.31 0 0 0 0 Frequency 0 0 3.58 0 0 2.93 2.19

0 0 0 0 0 0 Familiarity 0 0 0 0 0 560 0 0 0 0 0

14 9 9 7 5 Phon. Length 10 6 6 3 8 8 4 6 10 12 11 10

3 3 3 1 1 4 3 3 6 3 3 4 3 2 Syl. Length 3 2 3

5 7 15 7 5 7 9 10 Letter Letter Length 13 18 6 9 10 10 14 16 13

Word Count 2 2 1 2 2 1 1 1 1 1 1 2 1 3 2 2 1

Jones

chalice Word chalk chaps chameleon chandelier cellophane caviar Dion Celine centaur Celia Cruz Cruz Celia Cate Blanchett Cate Zeta Catherine Charles Durning Charles Lindbergh Charles chariot Gaulle de Charles Charles Manson Charles

104

112 17 17 17 17 17 First Syl. First Freq. 62 62 62 62 28 62 28 61 10 5 61

0 Phono. 0 0 0 0 9 3 0 ND 0 0 0 36 0 0 0 0 0

3.05 4.59 3.4 2.8 2.95 0 0 0 0 0 Frequency 0 0 0 0 0 0 0

0 469 0 0 0 0 Familiarity 0 0 0 0 0 588 0 0 0 0 0

10 8 11 13 8 Phon. Length 15 11 8 3 3 5 6 10 12 11 13 7

2 4 4 2 1 1 4 2 2 4 3 4 4 3 Syl. Length 4 4 5

6 13 9 12 6 10 4 6 Letter Letter Length 18 13 14 12 14 14 14 16 17

Word Count 3 2 2 2 2 2 2 1 1 1 2 2 1 2 2 2 1

checkerboard Word cheese chisel Cher cherub Charlton Heston Charlton Charlie Chaplin Charlie Theron Charlize Che Guevara Che Charlie Sheen Charlie Charles Martin Smith Martin Charles Schulz Charles Christine McVie Christine Reeve Christopher Chris O'Donnell Chris Rock Christopher Walken Christopher

105

Syl.

839 839 11 332 10 750 First First Freq. 2 9 35 1 0 34 11 58 58 326 839

0 Phono. 0 13 4 0 0 0 2 ND 5 0 9 0 0 0 24 28 7

3.51 3.53 0 0 0 4.02 3.78 0 3.61 2.61 Frequency 0 4.08 3.13 2.14 0 0 2.72

0 0 509 0 0 0 Familiarity 455 0 0 0 481 0 0 357 395 0 0

3 4 11 12 6 Phon. Length 8 5 4 6 9 8 9 4 6 4 12 3

1 2 1 3 2 4 2 3 1 2 1 3 4 2 Syl. Length 4 1 5

11 6 4 7 9 5 10 10 Letter Letter Length 13 4 7 4 13 14 7 5 14

Word Count 2 1 1 1 2 2 1 1 2 2 1 1 2 1 1 2 1

cockpit Word cockroach Powell Colin Coco Chanel Coco Firth Colin Cloris Leachman Cloris cleaver Eastwood Clint clove clef Chubby Checker clay compass conch collar comb Condaleeza Rice Condaleeza

106

1 60 51 104 571 231 First Syl. First Freq. 7 839 277 839 13 48 178 48 18 80 12

0 Phono. 51 10 7 9 0 6 7 ND 0 2 0 11 0 1 0 1 0

3.12 4.14 3.08 1.59 3.12 0 3.47 3.11 4.01 0 Frequency 3.47 3.38 2.5 4.15 3.84 3.46 3.05

0 441 0 0 0 0 Familiarity 425 521 433 478 0 531 506 0 0 416 0

5 5 4 6 6 Phon. Length 3 10 6 4 9 5 4 15 5 10 7 4

2 2 4 1 1 3 6 2 2 5 2 1 2 2 Syl. Length 1 2 2

6 14 6 7 5 8 8 5 Letter Letter Length 4 5 10 6 4 6 7 14 7

Word Count 1 1 1 1 1 1 1 1 1 1 3 1 1 1 2 1 1

crowbar Word crown crutch crucifix cruet cradle cornucopia crab crossbow cotton cone coral curtain Shepherd Cybill Cuba Gooding Jr. Cuba cuckoo cyclops

107

124 124 124 124 1 529 First Syl. First Freq. 87 127 151 31 127 127 127 127 127 8 124

0 Phono. 0 3 0 0 0 0 0 ND 0 0 0 0 12 0 0 0 0

0 0 3.29 0 0 0 0 0 0 0 Frequency 0 3.69 0 2.92 0 0 0

0 0 0 0 0 0 Familiarity 480 495 0 0 0 0 0 0 0 496 0

4 7 10 12 10 Phon. Length 9 8 13 9 4 10 8 16 14 9 9 12

5 5 3 4 4 1 6 4 2 4 2 3 5 3 Syl. Length 4 4 4

10 18 14 11 11 15 4 12 Letter Letter Length 11 6 9 8 11 14 14 10 11

Word Count 2 1 2 1 2 3 2 2 1 2 3 2 2 2 2 2 2

Lewis

Daniel Stern Daniel Word Danny Glover Bowie David dart Beckham David Daniel Day Daniel Dalai Lama Dalai Craig Daniel Daniel Radcliffe Daniel dandruff Cyndi Lauper Cyndi dagger David Tomlinson David Martin Dean David Clayton Thomas Clayton David Letterman David Debra Winger Debra

108

48 894 2965 42 151 151 First Syl. First Freq. 2965 894 264 27 529 320 320 320 320 320 16

2 Phono. 1 0 0 0 0 0 39 ND 0 0 0 0 0 31 0 5 0

0 0 0 4.02 0 0 2.47 3.61 3.99 0 Frequency 2.37 0 0 0 0 3.02 4.31

0 0 0 483 0 0 Familiarity 0 0 0 0 0 0 0 0 512 0 487

7 14 9 7 8 Phon. Length 7 7 6 9 7 3 9 13 3 10 6 5

2 1 4 3 4 4 7 1 3 3 5 4 4 4 Syl. Length 3 2 3

11 13 5 10 11 7 11 4 Letter Letter Length 8 10 9 16 9 8 4 11 6

Word Count 1 2 2 2 2 1 1 2 2 3 2 1 1 1 2 1 2

Dyke

Dian FosseyDian Word Keaton Diane Van Dick Diane Sawyer Diane dice diameter Demi Moore Demi Desi Arnaz diamond Denzel Washington Denzel decanter Wallace Dee dock Parton Dolly Diego Maradona Diego divot domino

109

26 0 320 418 418 933 First Syl. First Freq. 9 9 9 9 1 18 42 51 36 3 36

1 Phono. 0 0 0 0 0 0 5 ND 0 20 0 4 0 0 0 0 5

4.29 3.37 0 2.71 0 0 4 3.84 4.06 2.47 Frequency 0 0 0 0 0 3.75 0

478 0 0 465 0 349 Familiarity 0 0 0 415 425 439 0 0 0 0 0

14 8 7 3 6 Phon. Length 8 11 8 4 10 6 12 13 4 4 3 7

3 2 2 4 1 4 6 2 5 3 4 3 1 2 Syl. Length 3 3 2

13 17 8 6 4 7 13 8 Letter Letter Length 9 16 11 11 8 4 5 5 4

Word Count 2 2 2 2 2 1 2 1 2 2 3 1 1 1 1 1 1

Henley

dragon Word duel Dustin Hoffman Duke Ellington dumbbell dove Donald Trump Donald Doris Day Dr. Seuss Donna Summer Don Sutherland Donald eagle easel Dwight D. Eisenhower dynamite echo

110

54 933 933 54 54 54 First Syl. First Freq. 1302 0 0 933 0 54 54 933 933 933 1302

0 Phono. 1 0 0 0 0 0 0 ND 0 1 0 2 0 0 0 0 0

4.06 4.06 4.39 0 0 0 0 0 0 0 Frequency 3.45 0 0 0 0 3.79 0

0 0 0 0 0 0 Familiarity 0 0 0 564 601 459 0 0 0 0 0

7 7 10 4 11 Phon. Length 6 5 14 7 7 8 8 8 13 10 9 12

6 5 4 4 3 4 4 4 3 2 4 5 2 5 Syl. Length 2 6 4

10 10 15 11 8 16 8 10 Letter Letter Length 7 8 6 11 13 5 14 12 12

Word Count 1 2 2 2 3 1 2 1 1 2 2 2 2 2 2 2 1

electricity Word elephant Wiesel Elie elevator Manning Eli elbow Ed Koch Ed Poe Allan Edgar Eleanor Roosevelt Eleanor Eddie Murphy Eddie eclipse Harris Ed Ella Fitzgerald Ella Burstyn Ellen Elijah Wood Elijah Taylor Elizabeth Elliott Gould Elliott

111

71 933 27 418 933 933 First Syl. First Freq. 1302 54 54 397 397 397 1210 42 933 1302 933

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 2 0 0 0 0

4 2.61 3.01 0 0 3.12 2.5 0 0 0 Frequency 2.29 0 0 2.89 3.41 0 0

0 0 0 0 0 0 Familiarity 0 356 457 0 542 0 0 0 0 0 0

8 6 6 13 7 Phon. Length 7 11 10 10 5 11 12 8 7 10 7 7

3 3 6 5 4 3 4 4 3 4 3 3 6 3 Syl. Length 2 5 4

15 9 8 8 10 11 6 11 Letter Letter Length 8 9 12 6 7 13 10 11 8

Word Count 1 2 2 1 1 2 2 1 1 2 1 2 2 1 2 2 1

envelope Word equestrian Hemingway Ernest eraser Clapton Eric Emilio Estevez Emilio Elvis Presley Elvis emerald Enrico Fermi Enrico embryo ellipsis John Elton Ethan Hawke Ethan Longoria Eva escalator escargot Eva Peron Eva

112

100 22 1 19 6 5 First Syl. First Freq. 933 2 39 11 18 137 194 100 421 421 421

12 Phono. 0 0 0 6 7 0 0 ND 0 0 11 0 0 0 0 1 12

3.06 0 2.55 0 3.08 3.77 3.06 3.33 0 3.67 Frequency 3.73 0 4.23 0 4.48 0 3.82

496 0 499 0 401 434 Familiarity 0 0 564 0 0 0 523 562 485 0 0

12 6 4 8 4 Phon. Length 8 7 4 11 8 8 8 7 7 4 4 5

1 3 1 3 4 3 2 3 5 3 2 1 4 2 Syl. Length 4 2 1

8 10 5 5 11 7 8 10 Letter Letter Length 9 12 7 9 4 11 10 5 5

Word Count 1 2 1 2 1 2 1 1 1 1 1 1 2 1 1 1 2

lis

femur Word Castro Fidel flamingo filament rod fishing Faye Dunaway Faye factory farm feather Faith Hill Faith evolution McGregor Ewan fleur flock flashlight flask flute

113

172 1 61 93 6 221 First Syl. First Freq. 31 565 220 220 53 15 8 8 12 12 0

1 Phono. 45 0 4 2 0 0 8 ND 1 10 0 0 0 0 3 0 24

0 0 0 0 0 1.77 0 4.43 3.05 3.92 Frequency 2.47 2.84 4.28 3.19 3.84 4.15 0

0 0 0 0 0 530 Familiarity 402 0 469 494 0 0 338 513 0 0 0

7 5 6 4 17 Phon. Length 3 6 17 12 18 11 9 6 5 2 5 16

7 2 1 4 4 6 1 5 2 2 2 2 1 6 Syl. Length 1 5 2

10 5 16 18 12 16 18 14 Letter Letter Length 4 7 6 6 8 5 6 3 6

Word Count 1 1 1 1 1 1 2 3 3 2 1 1 2 2 1 1 2

Francis Ford Francis Coppola Word Frank Sinatra Kahlo Frida Franklin D. Roosevelt Franklin Mercury Freddie frame forest fountain Frances McDormand Frances fossil foal forceps funnel fur frond Batista Fulgencio garage

114

2 2 2 2 2 2 First Syl. First Freq. 125 125 221 221 87 34 1 1 1 61 2

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 2.74 3.78 0 0 2.96 0 0

0 0 0 0 0 0 Familiarity 509 0 0 0 0 0 0 0 0 0 0

6 9 6 9 7 Phon. Length 6 10 9 9 10 8 10 11 11 10 9 9

3 4 4 3 3 4 4 2 2 4 4 3 4 3 Syl. Length 2 3 3

12 14 13 9 10 11 10 11 Letter Letter Length 8 6 10 10 6 10 14 13 11

Word Count 1 1 2 2 1 2 2 2 2 3 2 2 2 2 2 2 2

Siskel

Gene Kelly Gene Word Gene C. Scott George Georg Solti Burns George Geena Davis Geena Gary Oldman gazebo Gene Hackman Gene Gary Sinise gargoyle garlic George Harrison George Kennedy George George Chakiris George Clooney George George Lucas George

115

9 3 3 3 3 19 First Syl. First Freq. 2 2 240 240 240 240 61 5 11 7 46

5 Phono. 0 0 0 0 9 1 0 ND 0 0 0 14 35 0 0 0 0

2.37 4.56 3.24 3.18 2.88 0 0 3.72 0 0 Frequency 0 0 0 0 0 0 0

477 409 0 0 0 0 Familiarity 0 0 0 0 0 505 0 0 0 0 381

14 14 11 12 4 Phon. Length 9 10 7 4 3 5 5 12 8 7 4 10

3 2 2 2 1 1 4 2 3 3 5 4 6 2 Syl. Length 3 4 1

7 13 13 6 5 8 4 7 Letter Letter Length 13 11 10 16 13 13 10 9 5

Word Count 2 3 2 2 2 2 1 1 1 1 2 2 1 2 2 1 1

geyser Word ghost glacier gill giraffe Geraldo Rivera Geraldo Gerald Ford Gerald Page Geraldine Geronimo Geraldine Chaplin Geraldine George Michael George Bush W. George Glenn Close Glenn Frey Glenn Glenda Jackson Glenda Campbell Glen globe

116

2 10 139 25 1 20 First Syl. First Freq. 27 32 32 5 1 26 5 106 26 26 46

0 Phono. 0 0 2 24 8 0 0 ND 0 47 28 0 0 0 0 0 0

3.33 0 0 2.87 2.29 0 3.31 3.66 3.34 0 Frequency 2.29 0 0 4 2.81 4.22 4.02

0 0 0 0 0 0 Familiarity 0 575 0 554 0 0 0 0 469 0 291

13 4 3 3 6 Phon. Length 10 10 3 9 8 7 6 8 7 10 6 7

3 2 4 2 4 3 3 3 6 5 1 1 1 2 Syl. Length 3 1 2

7 10 8 7 9 4 10 7 Letter Letter Length 12 13 12 5 5 4 10 11 7

Word Count 1 2 2 1 1 1 1 1 2 1 2 1 1 1 2 1 2

goggles Word Meir Golda gosling Goldie Hawn Goldie gondola goal Gloria Gaynor Gloria gnome goat glove glockenspiel Estefan Gloria greenhouse Gregory Peck Grace Kelly Grace graffiti grenade

117

1 183 183 7 183 356 First Syl. First Freq. 13 91 1 34 23 6 356 356 118 122 356

0 Phono. 0 0 0 48 5 0 11 ND 0 0 9 1 9 11 1 0 0

3.15 3.71 5.03 3.14 0 3.25 3.42 0 3.42 2.95 Frequency 0 2.93 0 2.29 4.08 0 4.1

0 0 0 430 421 423 Familiarity 0 0 440 0 0 0 0 0 515 0 0

7 8 3 8 5 Phon. Length 10 12 4 8 6 5 8 9 4 6 11 6

2 1 2 3 2 2 3 2 3 4 3 1 4 2 Syl. Length 3 2 4

9 9 7 7 9 6 6 6 Letter Letter Length 11 10 14 9 4 10 4 7 12

Word Count 2 1 2 1 1 2 1 1 1 2 1 1 1 1 1 2 1

hammock Word handcuffs Hank Aaron handle hanger Halle Berry Halle Gwyneth Paltrow hail hammer gyroscope Groucho Marx guillotine harp harpoon harmonica harness Harrison Ford

118

12 12 12 12 62 62 First Syl. First Freq. 356 356 356 #N/A 14 24 24 24 24 196 24

0 Phono. 0 0 #N/A 43 0 0 0 ND 4 0 0 0 0 2 0 0 0

0 0 4.2 3.98 0 0 0 0 0 0 Frequency 0 0 3.3 #N/A 3.87 0 0

0 0 0 0 0 0 Familiarity 0 #N/A 524 0 0 0 0 467 0 0 528

12 6 3 9 10 Phon. Length 10 5 9 16 9 6 10 9 10 9 8 12

5 4 4 4 7 4 3 2 5 2 2 1 3 4 Syl. Length 5 4 3

10 9 12 11 18 11 10 6 Letter Letter Length 12 12 7 8 4 9 11 13 10

Word Count 2 3 1 1 1 2 2 2 1 2 2 2 1 2 2 2 3

Helen Mirren Helen Word Carter Bonham Helena Henry Fonda helicopter helmet Helen Hunt Helen hatchet heel Helen Keller Helen hedgehog Harry Houdini Harry S. Truman Henry Thomas Hoover Herbert Henry Ford Henry Mancini Herman Cain Herman

119

23 63 20 37 43 63 First Syl. First Freq. 403 403 5 5 9 59 9 80 51 168 80

0 Phono. 0 0 0 4 59 22 0 ND 0 0 19 0 0 0 0 0 0

2.97 0 0 0 4.32 3.02 5.09 0 2.99 0 Frequency 1.59 2.25 0 0 2.77 4.09 3

0 498 419 0 548 0 Familiarity 0 0 0 0 489 0 0 0 386 0 0

5 13 4 5 2 Phon. Length 10 11 5 8 9 10 4 5 6 8 8 8

3 2 3 1 3 4 2 4 2 4 5 1 2 1 Syl. Length 3 1 2

4 9 8 3 11 4 11 11 Letter Letter Length 10 9 11 14 5 6 9 12 9

Word Count 1 2 2 2 1 1 1 1 2 1 1 1 2 1 2 2 2

hoe Word puncher hole horn Holly Hunter Holly Honus Wagner Hitler Hilary Swank Hilary hinge hive Hillary Clinton Hillary hieroglyph chair high hourglass Howard Hughes horseshoe hospital Hugh Grant

120

376 2968 386 386 386 386 First Syl. First Freq. 63 63 63 63 1 62 1302 2968 376 2968 2968

1 Phono. 0 0 0 0 2 0 0 ND 0 6 0 0 0 0 0 0 1

2.25 3.01 0 3.23 1.77 3.24 0 4.6 0 4.25 Frequency 0 0 0 0 0 2.87 2.5

507 0 0 0 0 555 Familiarity 0 0 0 437 0 0 0 0 0 0 485

8 10 9 5 7 Phon. Length 9 9 4 9 11 8 5 8 9 3 5 8

4 4 1 2 4 5 4 4 3 3 4 3 2 3 Syl. Length 3 2 2

5 8 9 7 9 5 12 8 Letter Letter Length 12 10 11 10 9 6 10 4 6

Word Count 2 2 2 2 2 1 1 1 2 1 1 2 1 2 1 1 1

incisor Word incubator ingot Indira Gandhi Indira infinity hurdle Hugh Jackman Hulk Hogan igloo Hugo Chavez Hugh Griffith Hugh Hefner Irene Papas Irene iron insomnia Cara Irene island

121

0 0 0 0 0 0 First Syl. First Freq. #N/A 22 22 22 48 48 48 48 48 48 0

0 Phono. #N/A 0 0 0 0 0 0 ND 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency #N/A 0 0 0 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 #N/A 0 0 0 0

9 15 7 8 9 Phon. Length 10 12 11 10 12 8 7 10 8 14 9 7

2 2 5 2 5 5 3 2 4 3 7 2 3 3 Syl. Length 5 4 3

9 12 9 10 13 13 14 10 Letter Letter Length 12 9 9 18 9 10 9 15 11

Word Count 3 3 4 3 2 2 2 2 2 2 2 2 2 2 2 2 2

Robinson

Jack Warden Jack Word Kennedy Jackie Caan James Jackie Jackie Brown James Jack Lemmon Jack J. R. Tolkien Jones Jack Jack Nicholson Jack J. Robert Oppenheimer J. Robert J Edgar Hoover J Edgar J. K. Rowling James Ford James Gandolfini James James Cameron James Dean James James Garner James

122

22 1 240 240 240 240 First Syl. First Freq. 0 0 22 2 2 2 74 74 5 22 74

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 27 0 0 0 0 0

0 3.92 0 2.69 0 0 0 0 0 0 Frequency 0 0 0 0 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 564 0 0 0 0 0

8 12 8 8 11 Phon. Length 11 8 9 3 12 7 4 9 11 10 9 13

6 5 5 2 1 4 3 3 3 3 5 3 3 4 Syl. Length 3 4 4

4 11 15 11 3 9 12 7 Letter Letter Length 13 11 9 13 9 11 14 11 11

Word Count 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 1

Z -

Janis Joplin Word jar Jay Jason Robards javelin Jane Goodall Jane Jamie Foxx Jamie Fonda Jane Janet Reno Janet Jane Alexander Jane James Hetfield James Taylor James Jennifer Warnes Jennifer Irons Jeremy Jeff Bridges Aniston Jennifer Jerry Herman

123

16 16 16 16 16 12 First Syl. First Freq. 240 240 240 240 240 240 240 6 16 16 16

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 0 0 0 0 0

4.04 0 0 0 0 0 0 0 0 0 Frequency 0 0 0 0 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 0 0

11 10 10 8 6 Phon. Length 7 10 10 7 7 12 8 9 7 9 9 9

4 3 4 3 2 2 4 4 4 4 4 4 4 3 Syl. Length 3 4 4

10 11 12 7 8 12 8 11 Letter Letter Length 9 13 13 11 12 10 9 12 11

Word Count 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2

jewelry Word Brown Jim Foxx Jimmie Jim Jones Jim Hendrix Jimi Jesse Owens Jerry Springer Jesse Jackson Jessica Lange Jessica Jerry Wexler Jerry Rice Jerry Seinfeld Jimmy Dean Jimmy Stewart Jimmy Jimmy Carter Jimmy Connors Jimmy Jodie Foster Jodie

124

9 9 9 9 9 9 First Syl. First Freq. 12 12 12 12 9 9 9 9 9 9 9

0 Phono. 0 0 0 0 0 0 0 0 0 ND 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 Familiarity 0 0 0

9 6 6 12 11 6 6 Phon. Length 9 7 8 9 10 8 11 10 11 10

4 3 4 3 3 3 2 4 4 3 4 3 5 5 2 Syl. Length 3 3

10 13 11 12 11 8 10 12 8 9 Letter Letter Length 11 10 8 8 13 13 12

2 Word Count 2 2 2 2 2 3 2 3 2 2 2 2 2 2 2 2

John F. Kennedy Word John Gielgud John Hurt John Lennon John Hartford Joe Montana John Amos John B. Anderson John Cazale Joe Pesci Joe Louis Joe DiMaggio John McEnroe John Steinbeck John Travolta John Wayne John Malkovich

125

6 6 6 6 63 1 First Syl. First Freq. 9 9 9 9 80 15 0 67 67 67 6

0 Phono. 0 0 0 0 0 0 0 ND 0 10 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 0 0 0 0 0 3.06 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 0 0

7 6 13 4 8 Phon. Length 8 7 7 10 11 9 11 9 10 13 8 10

4 4 7 5 4 5 4 4 3 3 2 7 1 3 Syl. Length 3 3 3

12 13 12 9 11 10 11 10 Letter Letter Length 11 11 10 9 13 5 13 13 10

Word Count 2 2 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2

Collins

Judi Dench Word Judy Roberts Julia Judy Garland Child Julia joust Johnny Depp Jose Feliciano Judd Hirsch Jon Voight Johnny Bench Johnny Cash Julie Christie Julie Iglesias Julio Julianne Moore Julianne Andrews Julie June Carter

126

38 38 38 38 38 38 First Syl. First Freq. 16 767 22 412 795 134 795 795 795 88 88

0 Phono. 6 0 0 0 0 0 0 ND 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 4.63 2.5 3.37 0 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 498 0 0 0 0

10 15 9 10 9 Phon. Length 4 7 12 10 9 9 11 9 11 9 10 10

4 4 3 5 4 4 4 3 4 3 6 4 3 4 Syl. Length 1 5 4

13 11 10 13 14 16 11 13 Letter Letter Length 4 12 8 17 10 11 12 10 11

Word Count 1 1 1 3 2 2 2 2 2 2 2 2 2 2 2 2 2

Jabbar

Bacon

Katharine Ross Katharine Word Turner Kathleen McGillis Kelly Keanu Reeves Keanu Richards Keith Kate Winslet Kate kangaroo Allen Karen Katharine Hepburn Katharine Kareem Abdul Kareem jury kaleidoscope Kevin Costner Kevin Kline Kevin Kenny Rogers Kevin Kevin Spacey Kevin

127

218 218 218 17 42 28 First Syl. First Freq. 3 0 89 124 326 326 18 66 218 7 45

0 Phono. 34 0 14 0 0 0 28 ND 52 0 0 30 0 1 0 5 0

3.02 3.97 0 3.31 0 0 2.95 0 3.29 4.28 Frequency 3.36 0 4.17 2.55 2.8 0 0

0 0 0 432 0 502 Familiarity 0 352 0 0 0 507 481 599 0 0 441

16 4 5 10 6 Phon. Length 3 2 8 4 12 6 7 7 3 6 7 4

2 1 2 3 2 3 3 2 7 1 2 3 4 2 Syl. Length 1 3 3

9 9 5 8 6 10 14 7 Letter Letter Length 4 19 4 7 5 10 5 7 9

Word Count 1 3 1 1 1 2 2 1 2 2 2 1 1 1 1 2 1

lacrosse Word ladder Bird Larry Lance Armstrong Lance lantern Kobe Bryant Kobe Bryant knee koala Kurt Cobain knocker kite Brandauer Klaus Maria latch laundry Larry King Larry lasso Laura Bush Laura

128

136 136 317 136 453 291 First Syl. First Freq. 28 66 218 28 136 136 136 136 317 317 1

0 Phono. 0 4 0 0 0 0 0 ND 0 0 0 0 0 33 0 0 0

0 0 0 3.67 0 0 3.19 4.36 0 0 Frequency 0 3.54 3.31 0 0 0 0

580 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 0 519

4 10 8 7 8 Phon. Length 11 7 13 5 11 4 11 16 9 8 7 8

3 4 3 4 2 4 8 1 2 3 4 3 2 3 Syl. Length 4 5 3

11 16 10 9 5 15 12 4 Letter Letter Length 14 4 8 15 10 8 10 8 7

Word Count 2 1 1 2 2 2 3 2 2 2 2 2 1 1 1 1 1

Trotsky

Lee Marvin Lee Word Lenin Leon Lenny Kravitz Lenny lens Lee Grant Lee lavender Rimes LeAnn Lee Harvey Oswald Harvey Lee Lawrence Taylor Lawrence Laurence Harvey Laurence lava Liam Neeson Liam Liberace Leonardo DiCaprio Leonardo leprechaun library

129

75 75 75 75 75 57 First Syl. First Freq. 75 453 453 453 291 453 291 66 9 162 75

0 1 Phono. 0 0 0 0 0 2 0 0 ND 0 29 7 1 1 0 0

3.16 3.87 2.95 2.61 0 0 0 4.05 0 0 Frequency 3.49 0 0 0 0 3.6 0

0 0 0 0 0 0 Familiarity 410 0 0 0 0 0 472 0 0 0 0

6 10 9 4 10 6 8 Phon. Length 9 14 4 6 4 7 9 15 5 11

4 4 8 3 2 2 2 2 2 4 3 5 4 2 5 3 Syl. Length 2

9 14 14 6 6 12 5 7 Letter Letter Length 10 10 9 17 12 6 16 11 7

Word Count 1 2 2 2 2 1 2 1 1 2 2 2 2 1 2 1 1

Word lobster locust llama loofah Lionel Richie Lionel Minnelli Liza Linda Hunt Linda litter Lindsey Buckingham Lindsey Lilia Skala Lilia lighthouse Louis Armstrong Louis Pavarotti Luciano Ball Lucille Lou Gehrig Lou Louise Fletcher Louise luggage

130

14 309 35 258 258 258 First Syl. First Freq. 11 3 384 384 520 48 52 520 520 40 520

0 Phono. 0 40 0 0 1 0 0 ND 0 0 0 0 59 1 0 0 0

2.9 2.33 2.78 3.09 0 1.89 0 0 0 0 Frequency 0 3.54 3.02 3.75 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 416 0

3 6 11 13 5 Phon. Length 14 6 9 6 3 7 7 6 15 11 11 10

4 6 5 3 3 1 3 3 1 3 3 4 5 2 Syl. Length 5 3 5

10 6 11 7 7 8 4 9 Letter Letter Length 14 4 7 7 12 15 18 14 16

Word Count 3 1 1 1 2 2 2 1 1 3 1 2 1 2 2 2 1

mammoth Word manatee Tung Mao Tse mane mannequin magnifying glass magnifying machete Johnson Magic Malcolm X Malcolm Madonna Lyndon B. JohnsonLyndon mace Margaret Rutherford Margaret Sanger Margaret maraca Anthony Marc Margaret Thatcher Margaret

131

520 6 138 2 151 3 First Syl. First Freq. 384 2 35 35 258 258 258 258 2 5 3

52 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 10 26 0 0 0

0 0 4.3 3.22 2.9 0 2.94 3.41 0 0 Frequency 0 0 0 0 0 0 0

434 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 0 427

12 9 12 18 12 Phon. Length 9 9 14 11 4 4 7 8 14 10 3 8

4 5 5 3 5 1 3 1 5 3 4 4 5 4 Syl. Length 4 5 1

7 9 9 14 14 14 4 4 Letter Letter Length 10 13 10 9 12 18 16 11 4

Word Count 2 2 2 2 2 4 2 2 1 1 2 2 1 1 2 1 3

Angelou

Marvin Hamlisch Marvin Word Moore Mary Tyler matador mask mast Martin Luther King Jr. Luther Martin Mark Hamill Brando Marlon Martin Scorsese Martin Mark Spitz Marie Curie Marie Monroe Marilyn Maximilian Schell Maximilian Maya Matt Damon Matt mausoleum maze

132

151 151 151 151 151 493 First Syl. First Freq. 428 13 13 111 428 428 428 145 151 151 151

0 Phono. 0 0 1 0 0 0 0 ND 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 Frequency 2.5 0 0 3.5 0 2.75 2.33

0 0 0 0 0 0 Familiarity 0 391 0 0 0 0 0 0 225 0 0

9 5 10 12 7 Phon. Length 7 12 8 8 11 12 11 11 11 10 10 11

4 4 3 4 3 4 4 5 3 4 2 3 5 4 Syl. Length 3 3 5

11 14 14 9 12 9 14 14 Letter Letter Length 9 9 13 7 11 13 13 13 13

Word Count 1 2 2 1 2 1 1 2 2 3 2 2 2 2 2 2 2

Douglas

Mia Farrow Mia Word Caine Michael J. Fox Michael Michael Douglas Michael Dukakis Michael metamorphosis Melvyn Melvyn Streep Meryl metronome mermaid megaphone Gibson Mel Michael Jordan Michael Phelps Michael Michael Jackson Michael Johnson Michael Michelle Obama Michelle

133

493 151 5 315 315 138 First Syl. First Freq. 493 493 493 151 151 1 151 151 493 151 493

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 0 9 24 31 1

0 1.77 2.71 4.12 3.29 2.71 3.63 0 0 2.37 Frequency 0 0 0 3.66 3.41 0 0

0 0 0 0 503 301 Familiarity 0 0 493 0 0 168 0 0 0 0 504

7 8 9 8 8 Phon. Length 9 10 13 7 7 5 7 4 13 7 12 4

1 6 3 3 3 3 2 2 3 4 3 3 3 3 Syl. Length 4 5 4

9 5 4 10 7 16 8 7 Letter Letter Length 16 10 12 10 10 9 14 7 13

Word Count 2 2 2 1 1 2 2 2 1 1 1 2 1 1 1 2 1

Gorbachev

Miles Davis Miles Word minaret mistletoe minotaur missile Mike Tyson Mike Mickey Mantle Mickey microscope Mikhail Mikhail microphone Michelle Pfeiffer Michelle Jagger Mick Monica Lewinsky Monica monocle miter mold Morgan Freeman

134

84 89 0 3 157 157 First Syl. First Freq. 16 309 40 210 309 3 14 309 1 1 1

0 Phono. 14 0 0 15 8 1 0 ND 2 0 0 0 0 0 11 0 0

3.21 0 0 0 3.34 4.08 0 0 0 0 Frequency 3.36 0 4.55 0 3.99 3.02 3.33

0 0 533 0 0 0 Familiarity 0 0 0 451 369 0 496 574 0 0 0

10 10 4 6 5 Phon. Length 3 6 6 9 13 9 5 5 13 10 11 11

3 5 4 3 3 5 2 4 5 2 5 2 2 2 Syl. Length 1 2 4

6 6 13 6 11 8 14 11 Letter Letter Length 4 12 8 11 5 5 13 11 12

Word Count 1 2 1 2 1 1 1 1 2 1 1 2 2 2 2 2 3

muzzle Word “Nat King” Cole nausea Natalie Portman Natalie Wood Natalie mural mountain mummy mushroom Muhammad Ali Muhammad moth Teresa Mother Nelson Mandela Cage Nicolas needle Armstrong Neil Nicole Kidman Nicole

135

486 419 1326 419 643 1 First Syl. First Freq. 5 7 1326 93 93 1326 3163 1326 68 1326 45

19 Phono. 19 28 0 0 0 1 0 ND 0 0 0 0 0 0 0 0 1

0 3.63 0 2.47 3.29 2.89 0 3.75 2.98 3.3 Frequency 3.34 3.84 0 0 3.29 0 0

477 0 0 358 0 0 Familiarity 500 0 370 0 0 550 304 0 0 0 0

3 13 7 11 9 Phon. Length 3 10 14 5 10 11 5 7 6 5 2 13

6 2 2 2 2 4 4 4 1 4 5 3 6 4 Syl. Length 1 7 1

7 7 13 10 5 14 12 11 Letter Letter Length 5 3 9 14 7 12 7 6 3

Word Count 1 1 3 2 1 2 2 2 2 1 1 1 1 3 1 1 1

Omar Sharif Omar Word onion orchard Oprah Winfrey optometrist Olivia Newton Olivia O.J. Simpson octopus Olympia Dukakis Olympia Octavia Spencer Octavia noose nun ostrich outlet origami Laden Bin Osama owl

136

618 79 264 618 618 8 First Syl. First Freq. 17 419 418 618 21 496 618 11 121 121 496

0 Phono. 6 0 18 0 0 0 0 ND 0 0 8 41 0 0 0 0 0

0 4.09 0 1.99 3.5 4.02 0 0 0 0 Frequency 3.49 0 0 3.58 2.85 3.05 2.77

0 0 429 0 0 0 Familiarity 0 0 387 0 0 566 453 0 301 0 0

9 5 6 6 7 Phon. Length 4 11 5 3 9 7 7 7 11 7 11 10

4 4 3 3 1 4 2 3 4 5 2 2 3 2 Syl. Length 2 2 4

9 8 12 8 3 7 11 7 Letter Letter Length 6 12 12 6 7 6 13 9 13

Word Count 1 2 2 1 1 1 1 2 2 1 1 2 1 2 2 2 1

Swayze

palm tree palm Word pan parachute Pancho Villa Pancho papyrus pagoda Pablo Picasso Pablo padlock palette paddle oyster OsbourneOzzy Patrick Patrick Duke Patty passport Neal Patricia Paul McCartney Paul

137

11 229 397 79 79 2 First Syl. First Freq. 8 8 8 8 496 17 184 184 229 229 14

0 Phono. 0 0 0 1 0 0 1 ND 0 31 0 0 0 52 0 1 0

3.16 0 0 3.73 2.69 0 3.25 3.46 2.74 0 Frequency 0 0 0 0 2.99 3.64 3.71

360 0 0 407 0 0 Familiarity 0 0 0 477 0 0 0 0 0 0 507

8 8 8 3 11 Phon. Length 13 10 5 14 10 3 6 4 8 12 7 7

3 3 5 2 6 4 2 1 3 3 3 4 1 5 Syl. Length 5 2 2

8 4 7 12 14 7 10 4 Letter Letter Length 17 10 12 9 8 4 8 12 7

Word Count 3 2 2 2 1 1 1 1 2 1 1 1 1 1 2 1 2

Glaser

pediatrician Word Pedro Almodóvar peephole Pedro Aznar peel pawn Paul Scofield Paul pavilion peacock Paul Simon Paul Paul Michael Michael Paul Newman Paul pendulum Cruz Penélope Pelé pelican penguin

138

85 16 16 29 396 396 First Syl. First Freq. 397 10 397 397 184 184 184 184 397 229 129

6 Phono. 20 0 10 0 0 0 19 ND 0 0 0 0 0 0 0 0 1

0 0 0 3.19 0 3.41 2.44 3.51 4.29 3.08 Frequency 4.39 0 3.14 2.8 0 0 0

0 0 489 535 0 409 Familiarity 0 209 0 0 0 0 613 0 0 0 0

7 5 8 8 8 Phon. Length 4 8 8 11 10 9 16 7 3 6 4 6

2 1 2 6 4 4 2 3 3 3 2 3 3 4 Syl. Length 2 4 2

20 9 8 11 11 9 13 10 Letter Letter Length 5 10 9 6 9 10 4 6 7

Word Count 1 2 1 1 2 2 2 2 2 3 1 1 1 1 1 1 2

Peter O'Toole Peter Word Clark Petula Hoffman Seymour Philip Peyton Manning Peyton pharmacist Peter Finch Peter periscope Falk Peter Peter Nero Peter pestle penny Faith Percy pipe piston pineapple pinecone pitcher

139

136 2 8 56 371 555 First Syl. First Freq. 33 53 53 32 17 23 37 69 418 9 159

1 Phono. 0 0 5 7 0 1 0 ND 0 2 0 2 0 0 0 0 2

0 3.09 2.44 2.82 2.69 3.3 0 1.89 0 3.19 Frequency 2.85 0 0 3.07 4.02 2.98 1.89

0 0 0 0 0 491 Familiarity 0 499 575 0 0 488 0 0 0 0 0

14 5 4 6 6 Phon. Length 7 6 8 5 10 9 6 7 12 7 9 5

1 5 3 3 2 4 2 3 6 2 2 1 2 2 Syl. Length 2 4 3

9 7 6 6 6 10 11 9 Letter Letter Length 9 14 6 6 4 7 13 9 10

Word Count 1 2 1 1 1 1 1 2 1 1 1 2 1 1 1 1 1

plunger Pol Pot poinsettia pitchfork platen pliers plug Placido Domingo Placido Word pollen potpourri pretzel Princess Diana propeller protractor pomegranate porcupine Prince

140

379 379 2253 379 217 217 First Syl. First Freq. 264 4 264 19 264 396 12 0 10 15 66

0 Phono. 17 1 0 4 0 1 3 ND 1 0 0 12 0 4 3 0 0

0 3.44 0 2.67 3.83 4.32 3.16 2.14 3.87 4.55 Frequency 3.29 4.04 3.74 0 3.87 3.6 0

0 515 523 486 0 0 Familiarity 0 0 486 386 0 409 521 382 0 0 368

7 5 5 7 16 Phon. Length 5 5 10 4 10 5 4 5 5 5 9 5

2 2 2 2 1 3 2 2 2 2 2 2 3 6 Syl. Length 2 4 5

5 6 7 16 5 12 11 6 Letter Letter Length 6 7 6 5 6 7 6 5 10

Word Count 1 1 1 1 1 1 2 2 2 1 1 1 1 1 1 1 1

Jones

pyramid Tarantino Quentin Queen Latifah Queen puddle puppet Putin puzzle pumpkin Word quill rabbi rabbit racket radar radiometer Quincy Quincy quiver raccoon

141

92 2 2253 4 70 3 First Syl. First Freq. 379 5 23 1 379 217 217 217 217 217 1660

0 Phono. 8 20 0 13 0 0 0 ND 43 0 0 0 0 0 0 0 0

0 0 0 0 0 0 3.92 0 0 0 Frequency 2.8 3.67 3.48 0 3.53 0 0

0 0 0 0 545 0 Familiarity 483 0 448 0 0 0 456 476 0 0 0

4 7 5 7 7 Phon. Length 5 3 7 10 8 12 8 11 7 11 11 11

5 3 5 3 4 3 4 5 1 1 2 2 3 4 Syl. Length 2 2 4

8 16 12 9 11 10 11 12 Letter Letter Length 6 4 4 12 6 9 11 14 11

Word Count 1 1 1 2 1 2 2 2 2 2 2 2 2 1 2 2 2

Ray Bolger Ray Liotta Ray Ray Charles Ray radish rake Fiennes Ralph rattle raft Word Mcnally Ray Baron Red Witherspoon Reese White Reggie Zellweger Renée Harrison Rex Raymond Burr Raymond McEntire Reba refrigerator

142

5 97 97 97 97 97 First Syl. First Freq. 173 4 4 4 4 4 7 7 69 1660 620

0 Phono. 0 0 0 0 0 25 0 ND 0 0 0 0 0 12 0 0 0

0 0 0 4.74 3.93 0 0 0 0 0 Frequency 2.88 0 0 0 0 0 2.78

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 400 0 0 0 565

10 8 13 9 10 Phon. Length 9 11 5 8 10 4 3 11 10 10 10 14

4 4 4 1 3 5 5 2 4 4 3 5 4 4 Syl. Length 4 2 4

4 12 15 11 10 8 10 5 Letter Letter Length 10 13 15 11 13 12 12 12 13

Word Count 1 2 2 2 3 2 1 2 2 1 3 2 1 3 2 2 2

Richard Pryor Richard Martin Ricky rickshaw rhinoceros Dreyfuss Richard Gere Richard M. Nixon Richard Richard Burton Richard Word Starr Ringo robe De Niro Robert Duvall Robert Goulet Robert Mitchum Robert Rita Moreno Rita river Downey Jr. Robert

143

97 97 97 97 97 182 First Syl. First Freq. 97 97 97 97 97 97 97 97 97 97 97

0 Phono. 0 0 0 0 0 0 0 ND 0 0 0 0 0 0 2 0 0

0 0 0 0 0 4.09 0 0 0 0 Frequency 0 0 0 0 0 0 0

0 0 0 0 0 0 Familiarity 0 0 0 0 0 0 0 0 0 0 0

12 13 7 10 10 Phon. Length 10 12 8 14 8 12 8 5 8 9 8 8

3 3 4 3 6 3 2 4 4 4 5 3 4 4 Syl. Length 4 3 3

11 5 10 12 15 10 9 13 Letter Letter Length 12 13 13 17 10 12 10 11 11

Word Count 2 2 2 3 2 2 2 2 2 2 1 2 2 2 2 2 2

Stewart Robert Wagner Robert Flack Roberta Robert Wise Robert Robert Palmer Robert Redford Robert Leonard Sean Robert Shaw Robert Robert Preston Robert Word Clemente Roberto Wright Robin robot Rod Waters Roger pin rolling Robin Gibb Robin Williams Robin Rod Steiger

144

302 188 367 367 367 29 First Syl. First Freq. 182 9 182 182 2253 11 150 302 150 150 12

0 Phono. 0 0 3 0 25 0 0 ND 0 0 0 0 11 21 0 11 0

3.59 0 3.5 3.15 0 0 3.89 0 0 0 Frequency 0 0 0 3.08 2.74 0 1.89

0 0 0 0 436 0 Familiarity 0 410 0 0 390 0 0 0 0 571 423

11 6 7 6 4 Phon. Length 13 9 5 11 4 3 8 10 8 11 12 5

2 3 4 3 6 2 4 1 4 4 3 3 3 2 Syl. Length 5 2 5

12 13 6 6 12 5 5 4 Letter Letter Length 13 12 9 6 7 11 10 14 12

Word Count 2 2 2 1 1 2 1 1 1 2 2 2 1 1 2 2 2

Word Polanski Roman Rosa Parks rosary Roy Scheider rudder Rudy Giuliani Crowe Russell Field Sally Ronald Reagan Ronald ruble ruler Hussein Saddam rung saddle Struthers Sally Dali Salvador rotunda

145

24 65 28 3 23 571 First Syl. First Freq. 367 4 4 367 105 105 367 50 50 7 112

8 Phono. 0 0 0 11 0 0 3 ND 0 0 0 12 20 32 0 0 0

2.8 2.65 3.06 5.6 3.23 2.52 3.25 3.83 3.67 0 Frequency 0 2.44 0 3.5 2.37 0 0

559 453 0 0 389 0 Familiarity 0 0 494 0 0 0 0 0 0 385 552

6 6 6 11 10 Phon. Length 14 14 15 4 3 2 8 6 5 15 5 7

2 1 5 3 1 1 2 1 2 5 3 2 4 4 Syl. Length 6 7 2

9 6 9 11 4 18 4 3 Letter Letter Length 12 6 14 7 6 13 5 17 8

Word Count 3 1 3 1 1 2 3 1 1 1 1 1 1 1 2 1 1

Word Davis Jr. Sammy Jackson L. Samuel samurai Bullock Sandra sarcophagus sari saxophone scarf sampan Parker Jessica Sarah sash scarab saw scarecrow Johansson Scarlett scissors sandal

146

13 13 13 79 79 79 First Syl. First Freq. 36 2 5 2 2 558 52 38 79 132 53

0 Phono. 0 2 0 0 0 0 5 ND 15 0 1 0 0 0 0 0 10

2.47 0 0 0 0 0 0 0 3.84 3.17 Frequency 3.43 3.36 2.52 0 0 0 2.25

0 0 0 528 0 398 Familiarity 350 0 0 0 0 0 342 0 0 0 0

5 8 6 13 6 Phon. Length 8 3 8 9 9 8 8 10 5 4 8 10

4 2 2 3 4 3 4 3 1 1 4 2 5 2 Syl. Length 3 2 3

12 15 13 8 14 7 11 10 Letter Letter Length 8 6 6 11 8 14 6 7 11

Word Count 1 1 1 2 2 2 1 3 2 2 2 1 2 2 1 1 3

Word O'Neal Shaquille Jones Shirley shovel Sharon Stone Crow Sheryl MacLaine Shirley Shirley Temple Shirley shutter shuttlecock scorpion scroll scythe Connery Sean Penn Sean Williams Serena shamrock sextant

147

1 1 8 18 185 13 First Syl. First Freq. 231 0 5 61 61 231 48 15 571 571 10

16 Phono. 0 0 0 0 0 15 0 ND 18 5 0 0 2 0 3 10 0

0 0 2.44 3.15 2.33 3.25 4.35 3.37 2.8 0 Frequency 3.18 0 2.74 0 0 2.98 3.33

513 0 489 0 501 0 Familiarity 0 0 0 0 0 0 0 0 0 0 0

11 11 11 4 10 Phon. Length 6 3 8 10 5 7 4 4 7 10 4 4

1 2 5 1 4 2 1 2 5 1 3 5 2 6 Syl. Length 2 3 1

5 5 5 13 11 10 5 9 Letter Letter Length 7 13 5 12 15 4 7 11 5

Word Count 1 2 1 2 2 1 1 3 1 1 1 1 1 1 2 1 2

Word Sissy Spacek sloop snorkel sitar slingshot snail snake Loren Sophia spade siamese Poitier Sidney sieve Freud Sigmund Weaver Sigourney silo O'Connor Sinéad silverware

148

47 47 118 82 118 74 First Syl. First Freq. 1 33 6 1 6 9 5 5 2 92 46

0 Phono. 5 0 2 0 5 0 1 ND 0 0 10 3 0 0 0 3 0

3.16 3.44 3.74 3.19 0 0 3.06 2.98 2.88 4.9 Frequency 3.07 3.05 0 3.02 0 3.83 3.02

333 0 0 0 0 548 Familiarity 407 0 0 538 0 0 0 0 433 511 0

7 6 6 5 4 Phon. Length 4 11 4 5 7 6 8 13 7 6 9 6

2 2 2 2 1 2 4 2 3 4 1 2 1 1 Syl. Length 1 1 3

8 14 6 4 5 5 8 6 Letter Letter Length 4 7 12 6 8 6 8 7 11

Word Count 1 1 2 1 2 1 1 1 1 2 2 1 1 1 1 1 1

Word squid Getz Stan starfish squirrel Stalin Kubrick Stanley staple station stethoscope spam spatula Tracy Spencer sphinx Lee Spike sponge spur spout

149

9 246 107 558 9 27 First Syl. First Freq. 1 1 1 1 1 1 16 114 6 5 11

26 Phono. 0 0 0 0 0 3 0 ND 0 7 0 6 3 7 0 0 13

2.41 3.22 2.25 2.99 3.39 3.85 0 4.42 2.94 3.83 Frequency 0 0 0 0 0 1.77 0

444 447 450 0 0 0 Familiarity 0 0 0 0 0 393 0 0 0 236 387

10 10 10 6 5 Phon. Length 10 13 4 5 5 4 5 8 9 4 4 13

5 3 1 1 1 1 3 1 3 4 4 4 1 2 Syl. Length 3 1 1

5 9 13 7 5 5 5 4 Letter Letter Length 11 12 15 11 12 5 10 4 5

Word Count 2 2 2 2 2 1 1 1 1 1 1 1 1 2 1 1 1

Word stork stump suspenders strop stub submarine Susan Sarandon swan sword Steve Martin Steve McQueen Steve Spielberg Steven Tyler Steven Wonder Stevie stilt stirrup Sting

150

1 1 156 19 4 50 First Syl. First Freq. 0 48 571 571 135 20 7 75 82 202 202

0 Phono. 0 0 0 21 0 0 0 ND 0 0 16 0 13 3 0 52 0

2.96 4.58 3.08 4.08 0 0 3.34 3.06 0 3.34 Frequency 0 0 2.99 3.29 3.62 2.95 4.41

0 0 0 0 0 481 Familiarity 0 0 417 0 0 0 0 431 511 0 0

14 6 4 8 9 Phon. Length 10 7 4 6 4 4 7 10 14 8 5 2

1 6 4 2 2 2 3 2 5 3 2 2 3 4 Syl. Length 4 1 2

7 11 3 9 6 4 5 6 Letter Letter Length 13 17 9 7 6 10 17 11 7

Word Count 2 2 1 1 1 1 1 1 1 2 2 2 1 1 1 1 1

Word target Cruz Ted Roosevelt Theodore tarot tattoo Williams Ted tee thermometer thermos Sydney Pollack Stallone Sylvester synagogue syringe tailor tambourine tarantula tank

151

0 156 9 134 17 17 First Syl. First Freq. 4 27 50 186 212 212 212 16 16 9 48

Phono. 1 31 0 13 2 0 0 ND 4 5 0 0 45 8 0 0 0 0

Frequency 2.47 4.12 0 3.26 4.65 4.27 0 0 0 0 2.69 4.09 0 2.97 0 4.52 0

0 0 0 291 571 Familiarity 0 513 0 0 546 397 547 0 559 0 0 0

Length 11 4 8 3 4 Phon. 6 5 5 2 8 9 8 6 10 8 7 4

2 1 3 4 3 4 2 1 Length 4 3 2 3 Syl. 2 2 1 3 2

7 7 16 5 3 5 10 8 0 4 4 10 10 5 6 9 10 Letter Letter Length

1 2 2 2 1 2 2 1 Word Count 1 1 2 1 1 1 2 2 1

Word thimble tire toga Tina Turner Tina tinsel Puente Tito toast Brady Tom Cruise Tom thunder Marshall Thurgood tiara tie tiger McGraw Tim Robbins Tim Tiger Woods Tiger

152

131 21 66 22 5 138 First Syl. First Freq. 17 17 17 17 2 134 134 6 148 148 148

Phono. 1 0 1 0 0 0 0 ND 0 0 0 0 10 0 0 7 0 0

Frequency 0 0 0 0 2.9 0 0 3.7 3.41 3.81 3.27 3 3.57 2.84 2.14 0 1.89

411 0 518 0 0 Familiarity 0 0 0 484 0 0 0 0 0 0 0 0

Length 8 12 4 7 9 Phon. 8 7 10 4 9 7 7 5 6 7 8 11

3 4 2 2 3 3 2 3 Length 3 4 4 1 Syl. 2 2 1 4 3

8 8 10 13 5 11 11 7 9 10 10 7 7 12 5 5 7 Letter Letter Length

2 2 1 1 1 1 1 1 Word Count 2 2 2 3 1 1 1 1 2

Selleck

Word Tom Hanks Tom tourniquet torpedo torso tractor Tommy Lee Jones Lee Tommy torch tornado tongs Braxton Toni Bennett Tony Tom Jones Tom Tom trampoline transom Tracy Chapman Tracy trapezoid

153

131 28 5 68 183 183 First Syl. First Freq. 100 134 134 4 11 26 19 0 52 101 131

Phono. 5 7 0 0 12 0 0 ND 0 0 1 0 1 0 1 0 0 9

Frequency 2.47 2.74 2.96 3.11 2.47 0 3.62 2.74 0 4.25 3.12 4.35 3.49 2.37 2.77 0 0

0 0 0 0 0 Familiarity 363 0 0 0 0 208 436 481 0 490 541 0

Length 6 11 7 4 3 Phon. 6 8 7 6 7 9 5 5 7 8 12 5

2 2 4 3 3 2 2 2 Length 2 2 5 3 Syl. 2 3 2 2 4

7 8 6 8 6 12 7 7 12 6 5 6 6 6 6 7 10 Letter Letter Length

2 1 1 2 1 1 1 1 Word Count 1 1 1 1 1 1 1 2 1

Word trellis turkey tunnel turban tuxedo trombone tsunami fork tuning trowel Capote Truman trumpet tricycle tripod udder ukulele Twiggy Ulysses Grant

154

34 21 144 112 21 73 First Syl. First Freq. 12 23 183 183 5 124 124 1 3 13 12

Phono. 18 0 0 1 0 0 0 ND 0 0 0 0 19 1 0 24 0 22

Frequency 0 3.88 3.4 2.37 3.29 3.76 0 3.59 2.94 3.75 4.27 0 3.45 0 3.52 3.68 3.06

0 461 0 0 468 Familiarity 0 487 452 0 0 0 511 0 369 0 472 457

Length 8 6 3 17 11 Phon. 8 7 9 2 13 6 4 3 6 7 7 6

4 3 3 2 2 1 1 7 Length 3 4 3 1 Syl. 4 3 1 5 3

10 8 7 8 3 6 15 7 10 17 13 4 8 6 4 4 6 Letter Letter Length

1 2 1 1 2 1 1 1 Word Count 2 1 1 1 1 2 1 1 1

Word Uma Thurman Uma Principal Victoria vest vise vineyard unicycle vase velcro urn vacuum Vanessa Redgrave umbrella unicorn Virginia Woolf Virginia volcano violin volleyball

155

7 19 5 107 34 49 First Syl. First Freq. 205 179 5 5 5 5 1 283 121 118 2

Phono. 9 0 10 2 4 0 0 ND 17 3 0 0 0 0 13 0 0 40

Frequency 4.12 4.36 3.06 0 0 0 3.49 0 0 6.48 3.76 2.8 4.12 4.19 0 2.77 0

0 0 476 505 0 Familiarity 506 0 0 0 0 0 558 0 0 381 550 510

Length 6 10 9 7 5 Phon. 4 5 12 9 4 7 3 3 3 12 11 4

2 1 4 1 3 1 1 3 Length 2 4 4 4 Syl. 2 2 4 1 4

6 6 6 14 13 12 4 14 14 11 7 4 4 5 5 12 11 Letter Letter Length

2 1 2 1 1 1 1 1 Word Count 1 1 1 2 2 1 2 2 1

Goldberg

Warren Beatty Warren weathervane wand Walter Payton Walter walrus Cronkite Walter Matthau Walter wallet Word waiter well wheelbarrow whip whistle Whoopi wheat whisk Houston Whitney

156

11 34 4 17 18 1 First Syl. First Freq. 1 149 25 25 25 25 25 25 25 110 110

Phono. 0 2 0 0 16 12 25 ND 39 3 0 0 0 0 0 0 0 0

Frequency 3.4 2.19 0 0 0 0 0 0 0 3.52 0 4.31 0 3.5 4.01 0 3.03

0 498 0 540 0 Familiarity 0 0 0 0 537 438 350 0 0 0 0 0

Length 7 12 7 4 3 Phon. 3 5 9 12 11 10 8 12 11 3 3 8

1 4 1 4 4 2 4 1 Length 2 4 4 3 Syl. 1 2 5 4 1

4 6 9 11 12 13 14 4 6 4 4 10 13 10 16 10 12 Letter Letter Length

2 2 2 2 1 2 1 2 Word Count 1 1 2 2 3 1 1 1 2

Willie Mays Willie Willie Nelson Willie William Shatner William William Holden William Will Smith Will Dafoe Willem H Macy William wicket Word wick windshield wolf Woodrow Wilson wool wreath Winston Churchill Winston Allen Woody worm

157

11 2 17 17 7 2 First Syl. First Freq. 17 8 0 0 14 15 15 7 15 15 38

Phono. 0 1 0 10 5 0 0 ND 0 24 0 0 0 0 0 11 0 0

Frequency 2.29 3.92 0 2.59 0 0 3.36 2.65 0 2.47 0 3.4 3.08 2.19 2.8 3.45 3.56

0 409 0 0 556 Familiarity 0 0 441 0 333 250 464 0 0 0 0 0

Length 4 6 3 5 4 Phon. 7 3 6 10 5 7 4 11 7 6 5 3

2 1 3 3 4 2 5 2 Length 2 3 3 1 Syl. 3 1 4 2 2

9 5 5 8 12 6 6 6 6 5 6 4 8 6 11 4 7 Letter Letter Length

3 1 1 2 1 1 1 2 Word Count 1 1 1 1 2 1 1 1 1

Yo Ma - yoke yarmulke Yo Yoko Ono zebra zither Word xylophone yacht Yanni Yasser Arafat yogurt yoyo Yuri Gagarin zeppelin zodiac zombie zipper

158

Appendix B. Stimulus Set List Assignment

Sublist 1 Sublist 2 Sublist 3 Sublist 4 Sublist 5 Sublist 6 Sublist 7 A B A B A B A B A B A B A B List 1 CN 8 17 17 16 16 11 14 13 24 14 13 14 11 18 (220) PN 22 15 13 16 14 21 17 18 9 17 20 17 22 13 List 2 CN 15 16 16 13 17 14 15 15 12 12 14 19 14 14 (221) PN 17 16 16 19 15 18 17 17 19 19 17 12 17 17 List 3 CN 12 13 14 13 19 16 19 17 11 14 14 14 15 17 (224) PN 20 19 18 15 13 16 13 15 21 18 18 18 17 15 List 4 CN 9 14 14 16 12 13 15 17 15 14 17 14 21 15 (217) PN 23 18 17 15 19 17 15 14 16 17 14 17 10 16 List 5 CN 18 13 15 12 13 16 14 12 15 13 13 19 14 16 (220) PN 14 19 17 19 19 16 18 20 15 18 18 11 17 15

Note. All numbers refer to word counts. Lists 1-5 were further divided into seven sublists. Each sublist contained both common and proper nouns. The “A” and “B” sets for each sublist represent the counterbalancing of cue type for all targets. Lexical items shown as a definition in the “A” set were presented as a picture in the “B” set. CN = common noun; PN = proper noun

159

Appendix C. Lexical retrieval task instructions

In this survey you will see pictures and read definitions, and you will try to name the word being represented. After each picture and definition, you will also be asked if you experienced a tip-of-the-tongue state. A tip-of-the-tongue state is the feeling of knowing a word but being unable to retrieve it. You will then respond to three rating scales that determine how frequently you have used the word in your lifetime and how familiar you are with the word. Frequency means how often you feel you use the word. Familiarity means how well you feel you know the word.

Be sure to answer the questions as quickly and as thoughtfully as possible.

160

References

Abrams, L., & Davis, D. K. (2016). The tip-of-the-tongue phenomenon: Who, what, and why. In H. H. Wright (Ed.), Cognition, language, and aging (pp. 13-53). Philadelphia, PA: John Benjamins Publishing Company. doi: 10.1075/z.200.02abr Abrams, L., Trunk, D. L., & Margolin, S. J. (2007). Resolving tip-of-the-tongue states in young and older adults: The role of phonology. In L. O. Randal (Ed.), Aging and the elderly: Psychology, sociology, and health (pp. 1-41). Hauppauge, NY: Nova Science Publishers, Inc. Astell, A. J., & Harley, T. A. (1996). Tip-of-the-tongue states and lexical access in dementia. Brain and Language, 54(2), 196–215. https://doi.org/10.1006/brln.1996.0071 Baayen, R.H., Piepenbrock, R., & Gulikers, L. (1995). The CELEX lexical database Philadelphia: University of Pennsylvania, Linguistic Data Consortium. Balota, D.A., Yap, M.J., Cortese, M.J., Hutchison, K.A., Kessler, B., Loftis, B., Neely, J.H., Nelson, D.L., Simpson, G.B., & Treiman, R. (2007). The English Lexicon Project. Behavior Research Methods, 39, 445-459. Beattie, G., & Coughlan, J. (1999). An experimental investigation of the role of iconic gestures in lexical access using the tip-of-the-tongue phenomenon. British Journal of Psychology, 90 (Pt 1), 35–56. https://doi.org/10.1348/000712699161251 Biedermann, B., Ruh, N., Nickels, L., & Coltheart, M. (2008). Information retrieval in tip of the tongue states: New data and methodological advances. Journal of Psycholinguistic Research, 37(3), 171–198. https://doi.org/10.1007/s10936-007-9065-8 Bowles, N. L., & Poon, L. W. (1985). Aging and retrieval of words in semantic memory. Journal of Gerontology, 40(1), 71–77. http://doi.org/10.1093/geronj/40.1.71 Brown, A. S., & Nix, L. A. (1996). Age-related changes in the tip-of-the-tongue experience. The American Journal of Psychology, 109(1), 79–91. Brown, A. S. (2012). The Tip of the Tongue State. New York, NY: Taylor & Francis. Brédart, S. (1993). Retrieval failures in face naming. Memory, 1(4), 351–366. https://doi.org/10.1080/09658219308258243 Brédart, S., Brennen, T., Delchambre, M., McNeill, A., & Burton, A.M. (2005). Naming very familiar people: When retrieving names is faster than retrieving semantic biographical information. British Journal of Psychology, 96(2), 205–214. https://doi.org/10.1348/000712605X38378 Brédart, S. (2017). The cognitive psychology and neuroscience of naming people. Neuroscience & Biobehavioral Reviews, 83, 145–154. https://doi.org/10.1016/j.neubiorev.2017.10.008 Brennen, T. (1993). The difficulty with recalling people’s names: The plausible phonology hypothesis. Memory (Hove, England), 1(4), 409–431. https://doi.org/10.1080/09658219308258246 Brennen, T., Baguley, T., Bright, J., & Bruce, V. (1990). Resolving semantically induced tip-of- the-tongue states for proper nouns. Memory & Cognition, 18(4), 339–347. https://doi.org/10.3758/BF03197123 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(5), 542–579. http://doi.org/10.1016/0749-596X(91)90026-G

161

Burke, D. M., & Shafto, M. A. (2004). Aging and Language Production. Current Directions in Psychological Science, 13(1), 21–24. http://doi.org/10.1111/j.0963- 7214.2004.01301006.x Burke DM, Shafto MA. Language and aging. In: Craik FIM, Salthouse TA, editors. The handbook of aging and cognition. New York: Psychology Press; 2008. Pp. 373–443. Button, K. S., Ioannidis, J. P. A., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S. J., & Munafò, M. R. (2013). Power failure: Why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365–376. https://doi.org/10.1038/nrn3475 Cohen, G., & Faulkner, D. (1986). Memory for proper names: Age differences in retrieval. British Journal of Developmental Psychology, 4(2), 187–197. https://doi.org/10.1111/j.2044-835X.1986.tb01010.x Cohen, G. (1990). Why is it difficult to put names to faces? British Journal of Psychology 81, 287-297. Cohen, G., & Burke, D. M. (1993). Memory for proper names: A review. Memory, 1(4), 249– 263. https://doi.org/10.1080/09658219308258237 Coltheart, M. (1981). The MRC psycholinguistic database. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 33A(4), 497–505. https://doi.org/10.1080/14640748108400805 Daffner, K. R. (2010). Promoting successful cognitive aging: a comprehensive review. Journal of Alzheimer’s Disease: JAD, 19(4), 1101–1122. Doi:10.3233/JAD-2010-1306 Dahlgren, D. J. (1998). Impact of knowledge and age on tip-of-the-tongue rates. Experimental Aging Research, 24(2), 139–153. https://doi.org/10.1080/036107398244283 Damasio, H., Grabowski, T. J., Tranel, D., Hichwa, R. D., & Damasio, A. R. (1996). A neural basis for lexical retrieval. Nature, 380(6574), 499–505. http://doi.org/10.1038/380499a0 Dell, G. S., Martin, N., & Schwartz, M.F. (2007). A case-series test of the interactive two-step model of lexical access: Predicting word repetition from picture naming. Journal of Memory and Language, 56(4), 490–520. https://doi.org/10.1016/j.jml.2006.05.007 Ecke, P. (2004). Words on the tip of the tongue: A study of lexical retrieval failures in Spanish- English bilinguals. Southwest Journal of Linguistics, 23, 1–31 Evrard, M. (2002). Ageing and Lexical Access to Common and Proper Names in Picture Naming. Brain and Language, 81(1–3), 174–179. Doi:10.1006/brln.2001.2515 Farrell, M.T., & Abrams, L. (2011). Tip-of-the-tongue states reveal age differences in the syllable frequency effect. Journal of Experimental Psychology. Learning, Memory, and Cognition, 37(1), 277-85. https://doi.org/10.1037/a0021328 Farrell, M. T. (2012). What’s in a Name? Predictors of Proper Name Retrieval Deficits in Old Age. (Publication No. 3569537) [Doctoral dissertation, University of Florida]. ProQuest Dissertations Publishing. Farrell, M. T., & Abrams, L. (2014). Picture–word interference reveals inhibitory effects of syllable frequency on lexical selection. Quarterly Journal of Experimental Psychology, 67(3), 525–541. https://doi.org/10.1080/17470218.2013.820763 Faust, M., Dimitrovsky, L., & Davidi, S. (1997). Naming difficulties in language-disabled children: Preliminary findings with the application of the tip-of-the-tongue paradigm. Journal of Speech, Language, and Hearing Research: JSLHR, 40(5), 1026–1036. https://doi.org/10.1044/jslhr.4005.1026

162

Festini, S. B., Hartley, A. A., Tauber, S. K., & Rhodes, M. G. (2013). Assigned value improves memory of proper names. Memory, 21(6), 657–667. http://doi.org/10.1080/09658211.2012.747613 Fogler, K. A., & James, L. E. (2007). Charlie brown versus snow white: The effects of descriptiveness on young and older adults’ retrieval of proper names. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 62(4), P201–P207. Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189–198. Foygel, D., & Dell, G. S. (2000). Models of impaired lexical access in speech production. Journal of Memory and Language, 43(2), 182–216. http://doi.org/10.1006/jmla.2000.2716 Frick-Horbury, D., & Guttentag, R. E. (1998). The effects of restricting hand gesture production on lexical retrieval and free recall. The American Journal of Psychology, 111(1), 43–62. https://doi.org/10.2307/1423536 Georgieff, N., Dominey, P. F., Michel, F., Marie-Cardine, M., & Dalery, J. (1998). Anomia in major depressive state. Psychiatry Research, 77(3), 197–208. https://doi.org/10.1016/s0165-1781(98)00008-0 Gianico-Relyea, J. L., & Altarriba, J. (2012). Word concreteness as a moderator of the tip-of-the- tongue effect. The Psychological Record, 62(4), 763–776. https://doi.org/10.1007/BF03395834 Gollan, T. H., & Silverberg, N. B. (2001). Tip-of-the-tongue states in Hebrew–English bilinguals. Bilingualism: Language and Cognition, 4(1), 63–83. https://doi.org/10.1017/S136672890100013X Gollan, T. H., & Brown, A. S. (2006). From tip-of-the-tongue (TOT) data to theoretical implications in two steps: When more TOTs means better retrieval. Journal of Experimental Psychology: General, 135(3), 462–483. Doi:10.1037/0096-3445.135.3.462 Gollan, T. H., Montoya, R. I., Cera, C., & Sandoval, T. C. (2008). More use almost always means a smaller frequency effect: Aging, bilingualism, and the weaker links hypothesis. Journal of Memory and Language, 58(3), 787–814. https://doi.org/10.1016/j.jml.2007.07.001 Goodglass H., & Stuss D.T. (1979). Naming to picture versus description in three aphasic subgroups. Cortex, 15, 199-211. Goral, M., Spiro, A., Albert, M. L., Obler, L. K., & Connor, L. T. (2007). Change in lexical retrieval skills in adulthood. The Mental Lexicon, 2(2), 215–240. Gorno-Tempini, M. L., & Price, C. J. (2001). Identification of famous faces and buildings. Brain, 124(10), 2087–2097. http://doi.org/10.1093/brain/124.10.2087 Hanley, J. (2011). Why are names of people associated with so many phonological retrieval failures?. Psychonomic Bulletin & Review, 18(3), 612-617. Doi:10.3758/s13423-011- 0082-0 Hanley, J. R., & Chapman, E. (2008). Partial knowledge in a tip of the tongue state about two and three-word proper names. Psychonomic Bulletin & Review, 15, 156–160. Harley, T. A., & Bown, H. E. (1998). What causes a tip-of-the-tongue state? Evidence for lexical neighbourhood effects in speech production. British Journal of Psychology, 89(1), 151– 174. https://doi.org/10.1111/j.2044-8295.1998.tb02677.x

163

Heine, M. K., Ober, B. A., & Shenaut, G. K. (1999). Naturally occurring and experimentally induced tip-of-the-tongue experiences in three adult age groups. Psychology and Aging, 14(3), 445–457. https://doi.org/10.1037//0882-7974.14.3.445 Hosmer, D.W. & Lemeshow, S. (1989). Applied Logistic Regression. New York: Wiley. IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp. James, L. E. (2006). Specific Effects of aging on proper name retrieval: Now you see them, now you don’t. The Journals of Gerontology: Series B, 61(3), P180–P183. https://doi.org/10.1093/geronb/61.3.P180 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. https://doi.org/10.1037/0278- 7393.26.6.1378 Jönsson, F. U., Tchekhova, A., Lönner, P., & Olsson, M. J. (2005). A metamemory perspective on odor naming and identification. Chemical Senses, 30(4), 353–365. https://doi.org/10.1093/chemse/bji030 Juncos-Rabadán, O., Facal, D., Rodríguez, M. S., & Pereiro, A. X. (2010). Lexical knowledge and lexical retrieval in ageing: Insights from a tip-of-the-tongue (TOT) study. Language and Cognitive Processes, 25(10), 1301–1334. https://doi.org/10.1080/01690961003589484 Kaplan, E. F., Goodglass, H., & Weintraub, S. (1983). The Boston Naming Test. Philadelphia: Lea & Febiger. Kavé, G., Knafo, A., & Gilboa, A. (2010). The rise and fall of word retrieval across the lifespan. Psychology and Aging, 25(3), 719–724. Doi:10.1037/a0018927 Kittredge, A. K., Dell, G. S., Verkuilen, J., & Schwartz, M. F. (2008). Where is the effect of frequency in word production? Insights from aphasic picture naming errors. Cognitive Neuropsychology, 25(4), 463. https://doi.org/10.1080/02643290701674851 Kohn, S. E., Wingfield, A., Menn, L., Goodglass, H., Gleason, J. B., & Hyde, M. (1987). Lexical retrieval: The tip-of-the-tongue phenomenon. Applied Psycholinguistics, 8(3), 245–266. https://doi.org/10.1017/S0142716400000291 Luce, P. A., & Pisoni, D. B. (1998). Recognizing spoken words: The neighborhood activation model. Ear and Hearing, 19(1), 1–36. Macizo, P., & Van Petten, C. (2006). Syllable frequency in lexical decision and naming of English words. Reading and Writing, 20(4), 295. https://doi.org/10.1007/s11145-006- 9032-z Marangolo, P., Rinaldi, M., Sabatini, U. (2004). Modality-specific naming deficit: Cognitive and neural mechanisms implicated in naming to definition. Neurocase, 10(4), 280-9. May, J. E., & Clayton, K. N. (1973). Imaginal processes during the attempt to recall names. Journal of Verbal Learning and Verbal Behavior, 12(6), 683–688. https://doi.org/10.1016/S0022-5371(73)80048-9 McWeeny, K. H., Young, A. W., Hay, D. C., & Ellis, A. W. (1987). Putting names to faces. British Journal of Psychology, 78(2), 143–149. Doi.org/10.1111/j.2044- 8295.1987.tb02235.x Menard S.W. (2001). Applied Logistic Regression Analysis. (2nd edition). SAGE, London. Meyer, A. S., & Bock, K. (1992). The tip-of-the-tongue phenomenon: Blocking or partial activation? Memory & Cognition, 20(6), 715–726. https://doi.org/10.3758/BF03202721

164

Meyer, A. S., Roelofs, A., & Levelt, W. J. M. (2003). Word length effects in object naming: The role of a response criterion. Journal of Memory and Language, 48(1), 131–147. https://doi.org/10.1016/S0749-596X(02)00509-0 Mortensen, L., Meyer, A. S., & Humphreys, G. W. (2006). Age-related effects on speech production: A review. Language and Cognitive Processes, 21(1-3), 238–290. http://doi.org/10.1080/01690960444000278 Nicholas, M., Obler, L., Albert, M., & Goodglass, H. (1985). Lexical Retrieval in Healthy Aging. Cortex, 21(4), 595–606. http://doi.org/10.1016/S0010-9452(58)80007-6 Oberle, S., & James, L. E. (2013). Semantically- and phonologically-related primes improve name retrieval in young and older adults. Language and Cognitive Processes, 28(9), 1378–1393. http://doi.org/10.1080/01690965.2012.685481 Ossher, L., Flegal, K. E., & Lustig, C. (2013). Everyday memory errors in older adults. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition, 20(2), 220–242. https://doi.org/10.1080/13825585.2012.690365 Peng, C.-Y. J., Lee, K. L., & Ingersoll, G. M. (2002). An introduction to logistic regression analysis and reporting. The Journal of Educational Research, 96(1), 3–14. https://doi.org/10.1080/00220670209598786 Perfect, T. J., & Hanley, J. R. (1992). The tip-of-the-tongue phenomenon: Do experimenter- presented interlopers have any effect? Cognition, 45(1), 55–75. https://doi.org/10.1016/0010-0277(92)90023-B Pine, K. J., Bird, H., & Kirk, E. (2007). The effects of prohibiting gestures on children’s lexical retrieval ability. Developmental Science, 10(6), 747–754. https://doi.org/10.1111/j.1467- 7687.2007.00610.x Pistono, A., Busigny, T., Jucla, M., Cabirol, A., Dinnat, A.-L., Pariente, J., & Barbeau, E. J. (2019). An analysis of famous person semantic memory in aging. Experimental Aging Research, 45(1), 74–93. https://doi.org/10.1080/0361073X.2018.1560118 Pyers, J. E., Gollan, T. H., & Emmorey, K. (2009). Bimodal bilinguals reveal the source of tip- of-the-tongue states. Cognition, 112(2), 323–329. https://doi.org/10.1016/j.cognition.2009.04.007 Ramscar, M., Hendrix, P., Shaoul, C., Milin, P., & Baayen, H. (2014). The myth of cognitive decline: Non-linear dynamics of lifelong learning. Topics in Cognitive Science, 6(1), 5– 42. https://doi.org/10.1111/tops.12078 Rastle, K. G., & Burke, D. M. (1996). Priming the tip of the tongue: Effects of prior processing on word retrieval in young and older adults. Journal of Memory and Language, 35(4), 585–605. https://doi.org/10.1006/jmla.1996.0031 Read, J. D., & Bruce, D. (1982). Longitudinal tracking of difficult memory retrievals. Cognitive Psychology, 14(2), 280–300. http://doi.org/10.1016/0010-0285(82)90011-1 Reason, J.T., & Lucas, D. (1984). Using cognitive diaries to investigate naturally occurring memory blocks. In J. E. Harris & P. E. Morris (Eds.), Everyday memory, actions and absent-mindedness (pp. 53–70). London: Academic Press. Redmond, S. M., Thompson, H. L., & Goldstein, S. (2011). Psycholinguistic profiling differentiates specific language impairment from typical development and from attention- deficit/hyperactivity disorder. Journal of Speech, Language, and Hearing Research: JSLHR, 54(1), 99–117. https://doi.org/10.1044/1092-4388(2010/10-0010)

165

Ross, L. A., & Olson, I. R. (2011). What’s unique about unique entities? An fMRI investigation of the semantics of famous faces and landmarks. Cerebral Cortex, bhr274. https://doi.org/10.1093/cercor/bhr274 Salthouse, T. A., & Mandell, A. R. (2013). Do age-related increases in tip-of-the-tongue experiences signify episodic memory impairments? Psychological Science, 24(12), 2489– 2497. https://doi.org/10.1177/0956797613495881 Semenza, C. (2006). Retrieval pathways for common and proper names. Cortex, 42(6), 884–891. Doi:10.1016/S0010-9452(08)70432-5 Semenza, C. (2011). Naming with proper names: the left temporal pole theory. Behavioural Neurology, 24(4), 277–284. Doi:10.3233/BEN-2011-0338 Shafto, M. A., Stamatakis, E. A., Tam, P. P., & Tyler, L. K. (2009). Word retrieval failures in old age: The relationship between structure and function. Journal of Cognitive Neuroscience, 22(7), 1530–1540. http://doi.org/10.1162/jocn.2009.21321 Shafto, M. A., James, L. E., Abrams, L., Tyler, L. K., & Cam-CAN. (2017). Age-related increases in verbal knowledge are not associated with word finding problems in the Cam- CAN cohort: What you know won’t hurt you. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 72(1), 100–106. https://doi.org/10.1093/geronb/gbw074 Storkel, H.L. (2018). Implementing evidence-based practice: Selecting treatment words to boost phonological learning. Language, Speech, and Hearing Services in Schools, 49(3), 482– 496. https://doi.org/10.1044/2017_LSHSS-17-0080 Tanaka-Ishii, K., & Terada, H. (2011). Word familiarity and frequency. Studia Linguistica, 65(1), 96–116. https://doi.org/10.1111/j.1467-9582.2010.01176.x Van Heuven, W. J. B., Mandera, P., Keuleers, E., and Brysbaert, M. (2014). Subtlex-UK: a new and improved word frequency database for British English. Q. J. Exp. Psychol. 67, 1176– 1190. doi: 10.1080/17470218.2013.850521 Vigliocco, G., Vinson, D. P., Martin, R. C., & Garrett, M. F. (1999). Is “count” and “mass” information available when the noun is not? An Investigation of Tip of the Tongue States and Anomia. Journal of Memory and Language, 40(4), 534–558. https://doi.org/10.1006/jmla.1998.2626 Villamar, M. F., Volz, M. S., Bikson, M., Datta, A., Dasilva, A. F., & Fregni, F. (2013). Technique and considerations in the use of 4x1 ring high-definition transcranial direct current stimulation (HD-tDCS). Journal of Visualized Experiments: JoVE, (77), e50309. http://doi.org/10.3791/50309 Vitevitch, M.S. (2003). The influence of sublexical and lexical representations on the processing of spoken words in English. Clinical Linguistics & Phonetics, 17(6), 487–499. https://doi.org/10.1080/0269920031000107541 Vitevitch, M. S., & Sommers, M. S. (2003). The facilitative influence of phonological similarity and neighborhood frequency in speech production in younger and older adults. Memory & Cognition, 31(4), 491–504. https://doi.org/10.3758/bf03196091 Waldron, E. J., Manzel, K., & Tranel, D. (2014). The left temporal pole is a heteromodal hub for retrieving proper names. Frontiers in Bioscience (Scholar Edition), 6, 50–57. Wechsler, D. (2008). Wechsler Adult Intelligence Scale (4th ed.). San Antonio, TX: The Psychological Corporation. 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.

166

Wong Adrian, Nyenhuis David, Black Sandra E., Law Lorraine S.N., Lo Eugene S.K., Kwan Pauline W.L., Au Lisa, Chan Anne Y.Y., Wong Lawrence K.S., Nasreddine Ziad, & Mok Vincent. (2015). Montreal cognitive assessment 5-minute protocol is a brief, valid, reliable, and feasible cognitive screen for telephone administration. Stroke, 46(4), 1059– 1064. https://doi.org/10.1161/STROKEAHA.114.007253 Yaniv, I., & Meyer, D. E. (1987). Activation and metacognition of inaccessible stored information: Potential bases for incubation effects in problem solving. Journal of Experimental Psychology. Learning, Memory, and Cognition, 13(2), 187–205. https://doi.org/10.1037//0278-7393.13.2.187 Zacks, R., & Hasher, L. (1997). Cognitive gerontology and attentional inhibition: A Reply to Burke and McDowd. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 52B(6), P274–P283. Doi:10.1093/geronb/52B.6.P274 Zamarian, L., Semenza, C., Trinka, E., Walser, G., Bonatti, E., Benke, T., and Delazer, M. (2008). Naming famous persons in patients with temporal lobe epilepsy, Epileptic Disorders, 10, 373–391. Zipf, G. K. (1935). The psychobiology of language. Boston, MA: Houghton Mifflin

167

168