Memory for General and Specific Words in Alzheimer's Disease

Material Information

Memory for General and Specific Words in Alzheimer's Disease
Argyros, Stephanie
Altmann, Lori ( Mentor )
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:


serial ( sobekcm )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.


This item has the following downloads:

Full Text

journal orf --5.earch

,,Oluiie ', issue 1 - SepEivem -ier .. Oc:oIer -i'u'i

Memory for General and Specific Words in Alzheimer's Disease

Stephanie Argyros


This study investigated people's memory for specific versus general words. This study is part of a larger study

that will examine the effects of using general versus specific words on the sentence production of individuals

with Alzheimer's disease (AD). Research has shown that individuals with AD use many general words in

spontaneous speech and have serious difficulties with sentence production when they must include very

specific words in their sentences (Altmann, 2004; Altmann et al., 2001). The underlying hypothesis of this work

is that using specific words puts a larger burden on memory than general words because their representations

include more semantic features than general words. These effects should be exaggerated in individuals

with Alzheimer's disease because they have impairments in both memory and the semantic system.

As a part of the larger study, we assessed 30 young adults (YA), 4 healthy older adults (HOA) and 2 adults with

AD by testing their memory of general and specific words in two tasks: one requiring just repetition of the words,

and one requiring recall and manipulation of the words. Both tasks are patterned after the Digit Span task

(Wechsler, 1987). Therefore, this study had two within-group variables (type of task: verbatim recall or recall

plus manipulation), and type of word (digits, general noun or specific noun) and one between-group variable

(group: YA, HOA and AD).

The findings of this study will help us determine whether specific words put more of a memory burden on

normal speakers than general words and allow us to test whether this effect is exaggerated by the

semantic impairment in Alzheimer's disease. It may also help us understand why "empty speech" is so common

in other people with semantic impairments, such as semantic dementia and Wernicke's aphasia.


Cognitive impairment in late life is a growing clinical problem, with Alzheimer's disease (AD) the most prevalent

of the progressive dementias. Dementia is characterized as a decrease in cognitive function. Memory disorders

are the most common and most disabling feature of neurodegenerative disease. It is important to understand

the symptoms of AD and to differentiate the diagnosis from other impairments such as semantic dementia

or Wernicke's aphasia because many symptoms may be similar. AD is a progressive neurological disease in which

all areas of the brain shrink and decline in function, leading to a slowly progressive decline in cognitive

performance. Consequently, symptoms subtly arise over time; judgment, reasoning, and problem-solving

become impaired, accompanied by changes in behavior and personality (Alzheimer's Association, n.d.).


AD causes a decline in memory, thinking, and language skills, and in the ability to read and write. At the

cellular level, two of the most important histological hallmarks of AD are neurofibrillary tangles and amyloid

plaques. Neurofibrillary tangles are twisted fibers of abnormal protein within neurons of the cerebral cortex. In

a brain affected by AD, amyloid plaques also form, which are an accumulation of protein fragments outside

the neurons. In an unaffected brain, these protein fragments are broken down, but in AD the mechanism

for metabolizing these proteins is ineffective.

It appears that the onset of cognitive decline correlates closely with the onset of neuronal loss in the

hippocampus and entorhinal cortex (Price et al., 2001). The hippocampus is a part of the brain located in the

medial temporal lobe that is responsible for the formation and consolidation of new memories. When damaged,

the ability to encode new memories and recall old memories is impaired, and disorientation can occur.

Moreover, Martin and Chao (2001) believe that the temporal lobe also plays a crucial role in the disease

process. These researchers state "patients with damage to the temporal lobes often have difficulty naming

objects and retrieving information about object-specific characteristics. This suggests that object-specific

information may be stored, at least in part, in the temporal lobes." The entorhinal cortex also contributes to

the encoding and consolidation of memories in the brain (Hodges, 1994). These two areas are the first regions in

the brain to suffer damage in AD.

As the disease progresses, the temporal and parietal association cortices become increasingly involved (Grossman

et al, 2003). These association cortices receive input from the primary sensory cortices of the brain which then

must integrate the information into representations from many modalities. The general impairment in naming due

to semantic impairment could be due to pathology of the region where the temporal, occipital, and parietal

lobes meet. Damage to these areas may lead to impairment in integrating different aspects of knowledge

and semantic information about words.

Semantic Representations

The meaning of a word is not only determined by the position of a word in a network of related words, but also

by smaller units of meaning called semantic features. As an example, if an individual hears the word "dog," there

are a plethora of features an individual may consider such as the dog "being furry," "having four legs,"

"playing frisbee," etc. Meanings of words intertwine because they share various combinations of semantic

features with other words. To illustrate, the meaning of the word "cat" decomposes into many features as

well, including "being furry" and "having four legs." Therefore, these two words share a subset of semantic

features. Some researchers believe that in individuals with AD, there is a progressive degradation of the

neural network where these shared features are represented (Grossman et al., 1993). Grossman et al. added

that "patients with Alzheimer disease encounter difficulty identifying and using diagnostic features that

contribute crucially to the semantic categorization of object descriptions" (1993) leading individuals with AD to

be particularly impaired when categorizing objects.

There has been a growing body of evidence from functional imaging studies pointing to differences in the

cortical representation for different concept categories (Martin, 1998; Martin et al., 1995) such as living versus

man-made objects or nouns versus verbs. For example, Pulvermiller (2005) defines action words as

"abstract semantic links between language elements and motor programs." He researched the

somatotopic organization of action words in the brain. The motor cortex in the brain, responsible for motor

functions, is organized in a way where the body is mapped out across the extent of the precentral gyrus (called

the motor homunculus). For example, control of the feet lies near the midline at the top of the gyrus, whereas

the lateral side of the gyrus controls the hands and face. Pulvermiller found that verbs dealing with foot

movements (e.g., kick) activate portions of the brain responsible for moving the feet. Likewise, verbs that deal

with mouth movement (e.g., lick) activate the portion of the brain that controls the mouth. Thus, action

verb meanings tap into the actual neural motor representations of those actions.

Semantic Impairments in AD

Researchers have used a number of tasks to investigate the semantic impairment in AD. Because a deficit in

word-finding ability has been characterized as an early sign in AD, researchers have often used a word-

picture matching task to measure comprehension and a verbal picture description task to elicit speech. A

semantic impairment has been described as broken down or degraded speech, based on the performance on

these tasks (word-picture matching and verbal picture description task) by individuals with AD (Bayles &

Tomoeda, 1983). The speech of individuals with AD is often described as lacking in content due to the overuse

of general words, but being relatively grammatical. In these tasks, individuals with AD usually produce few

content words and mostly empty elements such as pronouns, repetitions, and paraphasias. In general, studies

have shown this semantic impairment shown by individuals with AD is both due to a semantic knowledge

retrieval deficit and to a degradation of semantic memory in general (Croisile et al., 1996).

Semantic memory is described as "long-term memory containing knowledge of objects, facts and concepts as well

as words and their meaning" (Hodges, 1994) and is culturally shared rather than personal and is not

temporally specific, in comparison to episodic memory. Au, Chan, and Chiu (2003) assessed episodic and

semantic memory in a group of individuals with and without AD. The AD group had the largest impairment in

episodic memory, which was attributed to a deficit in encoding and storing new information. In addition,

results showed that semantic memory was affected progressively. According to Au et al., episodic

memory impairment is usually the primary feature of the disease. Additionally, these researchers confirmed

that impairment of naming ability is also one of the most common language deficits in individuals with AD.

Several researchers have found that AD may cause impairments of particular categories of words, such as

living things compared to man-made objects. In one of the first of these studies, Silveri et al. (1991)

hypothesized that a category-specific disorder could be present due to the structures of the brain that are affected

by AD in the early stages. Since these structures play a critical role in processing and storing information about

living things, they assessed 15 individuals with AD and 10 normal controls to test these categories using two

tasks. The naming task simply asked the participant to name the object in the picture; whereas the subjects

were required to say whether or not a picture and a word were related in the verbal association task. They found

the expected result that the individuals with AD performed more poorly on items from categories of living

things. Gonnerman et al. (1997) reported similar findings.

Tippett, Grossman, and Farah (1996) questioned whether the category effect that Silveri et al. (1991) had found

was a result of poor matching of stimuli on pertinent lexical variables. To confirm this, they first tested a group

of fourteen individuals with AD on the list of items used by Silveri et al. and found, as they had, a category

effect. They then tested the subjects on a second and third set of line drawings taken from Snodgrass

and Vanderwart (1980). The second set was composed of living (animals, fruits, vegetables and body parts) and

man-made objects matched pairwise for familiarity and word frequency. In the third set, 10 animals were

matched with 10 man-made objects for familiarity, word frequency, and visual complexity. Here, their test group

of individuals with AD did not demonstrate any category-specific deficit. Consequently, these researchers

attributed previous findings of the differences in performance on the two categories to other inter-

category differences.

Hodges et al. (1994) reported the performance of 22 individuals with AD on tasks of fluency (e.g., name as

many animals as possible in 60 seconds), picture naming, picture sorting according to three hierarchical levels,

word-picture matching, and verbal definitions. Pictures from the living category were chosen from the

subcategories of land animals, sea creatures, and birds, and the stimuli from the man-made categories

were household items, vehicles, and musical instruments. They found no category effect when analyzing the

group results or, in the case of the picture naming task, in looking for differences in individual cases.

To what extent is the functional-anatomic explanation tenable as a coherent explanation for the category effect

seen in AD? There has been a growing body of evidence from functional imaging studies pointing to differences in

the cortical representation for different concept categories (Martin, 1998; Martin et al., 1995). AD, though,

is distinguished by anatomic variability, yet the category effect appears present (and in the same direction) in

the large majority of individual cases. This would seem to call into question any explanation predicated on

the anatomic distribution of cortical degeneration in AD. Many studies have reviewed the large set of variables

that can impact on the effect, and a range of theories have attempted to explain it. Many researchers now

believe that the category effects in AD may well be an emergent property of the organization of semantic

memory itself.

Studies contrasting object and verb knowledge have produced contradictory results with some reporting better

action naming and some superior object naming. Cappa et al. (1998) compared 19 individuals with AD to

10 individuals with frontal-temporal dementia (FTD) on a task of naming pictures for objects (both living and

man-made) and pictures depicting actions such as washing or pulling. They found both groups to be worse at

naming actions, rather than objects, but the FTD subjects showed a greater category difference.

Robinson, Grossman, White-Devine, and D'Esposito (1996) also found that 17 out of 20 individuals with AD

were worse at naming pictures of actions than pictures of objects. They were also worse at matching a verb in

print to its appropriate illustration out of an array of four than performing the same task with nouns. Bushell

and Martin (1997) also found abnormal semantic priming for verbs but not for nouns in individuals with AD.

They found that semantic priming facilitated the performance of elderly normal controls in reading concrete

nouns and motion verbs, but that individuals with AD were facilitated by semantic priming only in reading

concrete nouns. Neither group of subjects showed priming by semantic associates for either abstract words or

non-motion verbs. The prime-target relationship, however, appears to have been different for the four types

of words. Primes for abstract and non-motion verbs were predominately synonyms of the target (e.g.,

freedom-liberty; instruct-educate), whereas for the concrete nouns they often were not (e.g., gold-silver; tiger-

lion) (Bushell & Martin, 1997). Perhaps synonyms are poorer semantic primes, which would partly explain the results.

In contrast, Williamson et al. (1998) found a specific deficit in naming nouns compared to verbs. They suggest

that lexical retrieval is also an important functional impairment in AD. The pattern of loss for action naming

versus object naming might well reflect the different "neural substrates" for lexical retrieval of nouns versus

actions. Impaired lexical retrieval for objects would be due to lateral and inferior temporal lobe pathology,

whereas impaired lexical retrieval for actions would involve damage to premotor or pre-frontal cortical areas or

both, which are relatively spared in AD. Williamson et al. suggest that the category effect with specific impairment

of object naming might, therefore, be at the level of lexical-phonological retrieval and not linguistic semantics.

One possibility that may explain the different performance with nouns and verbs was suggested by Fung et

al. (2001). Fung et al. contended that it was problematic to compare naming an action depicted by a static

picture with naming the picture of an object. Whereas the object is fully represented in its static

visual representation, identifying and naming an action requires a further level of abstraction across time.

Moreover, they argued that in some cases recognition of the action could be facilitated by first recognizing

objects from the scene (e.g., we know the man is skating because he is wearing skates), which would be to

the advantage of normal subjects but not necessarily to the individuals with AD. In this study, individuals with

AD performed better on the man-made categories than on the living categories, and better on the action verbs

than on the abstract nouns or the living categories. Accuracy for abstract nouns was not significantly different

from living nouns. The results from both tasks indicated, first, that knowledge of action verbs was not more

impaired than knowledge of objects in AD. Second, the convergence of results from tasks using different

stimulus types (i.e., pictures and words) and different response type (i.e., naming versus forced choice

judgments) calls into question explanations of category effects that are based primarily on visual-perceptual

or lexical-phonological deficits.

Individuals with AD also show impairment when using function words. Function words may include pronouns (e.

g., he, her), prepositions (e.g., on, beneath), auxiliary verbs (e.g., be, do, have), or conjunctions (e.g., and,

but, nor). Function words have little lexical meaning, but serve a purpose to express grammatical

relationships between other words within a sentence. Function words, also called "closed-class" words, usually

are activated easily, because they are very high frequency (Bock, 1989) and their semantic representations consist

of only a few distinguishing features. Closed-class words also consist of a relatively small set of items, which do

not offer possibility for expansion as much as "open-class" words. Open-class words consist of nouns, verbs,

and adjectives. Altmann et al. (2001) challenged the idea that, in individuals with AD, a semantic impairment

usually leads to impairment in open-class words, but closed-class word use is relatively preserved. To address

the ambiguities of previous research which stated that during the early stages of AD, morphosyntax is

preserved, these researchers examined the frequency of errors in morphosyntax and lexical choice among

these individuals. Altmann et al. found that individuals with AD made errors in all aspects of language, including

both open-class and closed-class words, during spontaneous speech.

Spontaneous Speech

AD has been found to have profound effects on spontaneous speech. Research has shown that the speech

production of individuals with AD is severely impaired when they are unable to choose the words they must include

in a sentence; however, fewer errors are made when individuals are required to produce speech themselves

without constraint.

Early analyses of spontaneous speech in individuals with AD reported that syntax was relatively spared.

Although their spontaneous speech is marked by impaired lexical access and semantic emptiness, it was

nonetheless syntactically well-formed (Hier, Hagenlocker, & Shindler, 1985). Kempler, Curtis and Jackson (1987)

also found that individuals with AD had many lexical errors but fewer syntactic errors. Moreover, the frequency

with which they used various syntactic constructions was the same as normal subjects, as was the length

and grammatical complexity of their utterances. In another assessment of grammatical knowledge, individuals

with AD were asked to correct errors in spoken speech as a means of assessing comprehension; they failed to

correct semantic errors but managed to correct errors of syntax (Bayles & Boone, 1982). Although these

findings suggest that syntax may be relatively spared, others have suggested the opposite.

Bates et al. (1995) conducted a study using a constrained sentence production task and found that the

participants with AD had difficulty producing complex sentences. Results also showed that the participants

had trouble producing grammatical sentences in the most difficult conditions, which was attributed to a lexical

deficit. Altmann et al. (2001) and Bates et al. agree that the semantic impairments of individuals with AD can lead

to difficulties with the general production of grammatical sentences.

Altmann et al. (2001) conducted two experiments, one measuring the frequency of errors in spontaneous speech

and another measuring errors in a constrained speech production task. In the spontaneous speech, results

showed that individuals with AD produced the same type of errors compared to healthy older adults; moreover,

they made more errors in all four categories measured: open-class words, closed-class words, pronouns,

and morphosyntax. These findings showed that morphosyntax is not preserved in the spontaneous speech

of individuals with AD. The second experiment tested the hypotheses that individuals with AD have impairment

when they must include specific nouns and verbs in a sentence. The constrained sentence production task

was designed to elicit passive sentence production through the manipulation of the verb type and the order of

nouns presented. The participants were given an index card containing a verb and two common names (e.g.,

Tommy or Susan) or a verb, a common name, and an inanimate noun (e.g., ball, toy, movie) and were asked

to create a grammatical sentence that included all three words.

Results showed that individuals with AD were significantly impaired when asked to provide the

necessary grammatical closed-class words that were needed to make a grammatical sentence. Furthermore,

while both groups produced morphosyntactic errors, individuals with AD produced more. In addition, Altmann et

al. (2001) reported that individuals with AD produced more grammatical errors in the constrained

sentence production task than they had in the spontaneous speech task. Therefore, AD affects not only production

of semantically-represented words like nouns and verbs, but also the production of grammatical closed-class

words (Altmann, 2004).

Altmann et al. (2001) concluded that speech in individuals with AD is highly impaired in comparison to healthy

older adults when they are required to include specific nouns and verbs in their sentence production. When

specific words had to be included in a sentence, omission or substitution of one of these words occurred. There

are two hypotheses that may account for why this occurs. First, individuals may have more difficulty activating

and maintaining the activation of the richer semantic representations of specific words in memory; whereas,

they may have an easier time activating and holding the sparser semantic representation of general words.

An alternative hypothesis is that individuals with AD have a severe impairment activating and maintaining

the activation of any word not of their choosing; in this case, activating and recalling general and specific

words would be equally difficult for them. To test these hypotheses, we used a verbatim memory and

verbatim memory plus manipulation task using digits, general words, and specific words. We predicted that

digits would be the easiest to repeat verbatim and to manipulate than doing so with words in a word span task.

We also predicted that subjects would have better memory for general words compared to specific words, and

that this difference would be exaggerated in a recall plus manipulation working memory task.



Participants included 30 young adults between the ages of 17 and 35 (mean = 22, SD = 3.4) and years of

education ranging from 12 to 21 (mean = 15, SD = 1.6). Twenty-seven females and 3 males participated in

our experiment.

As a part of a larger study, we also assessed 4 healthy older adults (3 females and 1 male) between the ages of

76 and 84 (mean= 80, SD= 3.9) and years of education ranging from 13 to 18 (mean= 15, SD= 2.6) and

2 individuals (1 male and 1 female) with AD of ages 70 and 78 respectively (mean= 74, SD = 5.7) with 12 and

18 years of education (mean = 15, SD = 4.2) on similar measures during three separate visits.


Forty-five pairs of semantically related, general and specific words were compiled, where the general word was

a superordinate term for the specific word (e.g., nut-acorn, season-autumn, furniture-sofa). General and

specific words formed separate lists and were matched groupwise for word frequency and familiarity

(where available). For each of the four lists, the number of syllables was matched at each memory list length (e.

g., for list length of four words, all lists had seven syllables). However, the words appeared in different

positions across each list. Some words appeared twice in each list, but all words were used once before any

were repeated. Thus, repetitions began in the seven-word list. Only 5 of the 30 subjects reached that level in any

test version.


Over the course of the study, each participant was asked to complete three memory tasks. The order of

presentation of specific and general words in the verbatim recall and recall plus manipulation word lists

were counterbalanced across subjects.

First, the participants were given the Wechsler Digit Span task which was used to investigate verbal working

memory for digits. In the first memory task, the Digit Span Forward was administered by the experimenter,

who gave these directions, "I am going to say some digits. Listen carefully and when I am through, say them

right after me." The experimenter began by saying three digits aloud (eg., 3-8-5), and the participant repeated

the list verbatim. The list of digits increased and testing was discontinued if the participant failed on two

consecutive trials at a given test length. Then, the Digit Span Backward was administered. The

experimenter explained the directions by saying "Now I am going to say some more digits, but this time when I

stop I want you to say them backwards. If I say 7-1-9, you would say 9-1-7." Participants repeated the digits in

the reverse order in which they were presented, and, similarly, the test was discontinued after failure on both sets

of a given length.

The next two tasks were patterned after the Digit Span task and tested the participant's verbal working memory

for general and specific words. In the verbatim recall-only version of the general word span task,

participants repeated increasingly long lists of general words. For example, the experimenter would begin by

saying "animal, season, bag." If the participant repeated the series correctly in two different trials, the lists

would become longer by one word. As in the Digit Span tasks, testing was discontinued if a person failed on

two consecutive trials at a given list length. In the recall plus manipulation version, participants were asked to

repeat the general words that were presented to them in alphabetical order. If the experimenter, read

"sport, vegetable, meat," the appropriate response from the participant would be "meat, sport, vegetable." When

at least one list was reproduced correctly, the experimenter continued testing at the next longer list.

The experimenter discontinued testing if the participant failed in both trials of the same list length.

The verbatim recall and the recall plus manipulation tasks were also given using a list of specific words. In both

word span tasks, the verbatim recall was always given first. However, the order in which the general word

and specific word tests were given was counterbalanced across subjects.


Results were analyzed using a (3) word type by (2) levels of processing repeated measures ANOVA. The three

word types were digits, general words, and specific words. The two levels of processing were a verbatim recall and

a recall plus manipulation task. In the analysis examining the effects of word type and task type, there was

a significant effect of word type, F(2, 28) = 37.142, p<0.001, r72 =.73. The YA participants remembered more

digits than either general or specific words (both p <0.001). Moreover, participants were better at

remembering general words than specific words but this did not reach significance (p = 0.10). Furthermore,

there was no effect of task type. However, there was a significant interaction between word type and task type, F

(2, 28) = 8.958, p < .001, r/2 =.39. As shown in Figure 1, participants remembered more digits in the verbatim

recall task than in the recall plus manipulation task, as expected. However, in both tasks using words,

participants remembered more words in the recall plus manipulation task than in the verbatim recall task, which

was unexpected.

o Ldt.'c. I .T I '
a Real plus maniulana


2 . / --

Digris Genwal Wheoft SpWieW ord
Word Type

Figure 1. The number of digits, general, and specific words remembered in the verbatim recall and

the recall plus manipulation tasks.

Preliminary results from the HOA and AD group showed similar results. Both groups remembered more digits

than general and specific words in the verbal working memory tasks, as predicted. As shown in Table 1, the

HOA group remembered more words in the recall plus manipulation task than in the verbatim recall task,

which directly corresponds with the results of the YA group. The 2 participants with AD remembered more

general words in the recall plus manipulation task, but the same amount of specific words in both the verbatim

recall and the recall plus manipulation task.

Table 1.
Word Type mean Results

Digits General Words Specific Words
Verbatim Manipulation Verbatim Manipulation Verbatim Manipulation

YA 10.1 (2.1) 8.7(2.8) 6.6(1.5) 7.3 (1.4) 6.3 (1.6) 7.0(1.3)

HOA 7.0 (1.4) 7.8 (3.3) 5.5 (1.3) 5.3 (.96) 5.5 (1.7) 6.8 (1.3)

AD 6.5 (.71) 4.0 (1.4) 3.0 (0) 3.5 (.71) 3.5 (.71) 3.5 (.71)


We designed this study based on findings that individuals with AD use many general words in spontaneous

speech but have serious difficulties with sentence production when they must include very specific words in

their sentences (Altmann, 2004; Altmann et al., 2001). The purpose of this study was to test if specific words put

a larger burden on memory than general words, because their representations include more semantic features

than general words. We predicted that words with richer semantic representations would be more difficult to

recall. Specifically, we predicted that digits would be the easiest to recall in comparison to general words and

specific words would be most difficult. Our second prediction was that the effect would be exaggerated if

the individual had to recall and manipulate words.

As expected, our results show differences between digit and word recall. We found a continuum in how many

items were remembered; more digits were recalled than general words, and more general words were recalled

than specific words. Unexpectedly, participants found it easier to recall and manipulate words than to simply

store them in memory.

We also found effects in the different types of manipulation, specifically, the Digit Span backward and the recall

plus manipulation task, using both general and specific words. The ordering of digits in the Digit Span backward

is arbitrary and has no intrinsic order. In the recall plus manipulation task, participants may have used the

alphabet in a process termed "scaffolding," to structure their memory of lists. Scaffolding is defined as a

symbol system, which holds detailed representations internally in long-term memory. There are various types

of memory scaffolding, such as rhymes or codes, which individuals may use to augment the memory (http:// Anecdotally, several YA participants reported that they were actually alphabetizing the words

as the experimenter presenting them. This process would qualify as the use of a scaffolding strategy. This

temporary framework may explain why participants remembered more words in the recall plus manipulation

task than in the verbatim recall task.

These findings suggest that investigation of the interaction between word type (digits, general, and specific

words) and processing type (verbatim recall and recall plus manipulation) needs to be conducted. In particular,

all word types should be tested in all conditions: verbatim recall, backward recall, and ordering. Therefore,

an ordering task, such as Digit Ordering, could be added to the digits list. Similarly, general and specific

word memory needs to be tested using a method similar to the Digit Span backward, in which word lists would

be recalled in the reverse order of presentation. Moreover, the ordering of task types could also be

counterbalanced to further examine why participants found it easier to recall and manipulate words rather than

just recalling them.

This study adds to our knowledge about the relationship between the semantic system and memory. The

verbal working memory tasks suggest that individuals may ultimately find it easier to remember items which hold

a sparser semantic representation, such as digits and general words, than to recall specific words which is

more taxing due to their many semantic features. Finally, this study has given us a solid foundation for future

studies examining effects of task type and word type in healthy individuals and those with semantic impairment,

such as AD. In future studies, we predict individuals will perform better in the Digit Ordering task than in the

Digit Span forward or backward. Moreover, we predict there will be an exaggerated effect of word type in the

word backwards tasks. Similar to our previous hypothesis and findings, we believe there will be a continuum in

how many items are remembered: digits will be the easiest to recall and manipulate, followed by general and

then specific words.


1. Altmann, L. J. P. (2004). Constrained sentence production in probable Alzheimer disease. Applied

Psycholinguistics, 25, 145-173.

2. Altmann, L. J. P., Kempler, D., & Andersen, E. S. (2001). Speech errors in Alzheimer's disease:

Reevaluating morphosyntactic preservation. Journal of Speech, Language, & Hearing Research, 44(5), 1069-1082.

3. Alzheimer's Association. (n.d.). What is Alzheimer's?Retrieved January 29, 2007, from www.alz.orq.

4. Au, A., Chan, A., & Chiu, H. (2003). Conceptual organization in Alzheimer's dementia. Journal of Clinical

and Experimental Neuropsychology, 25(6), 737-750.

5. Bates, E. (1995). Production of complex syntax in normal aging in Alzheimer disease. Language and

Cognitive Processes, 10, 487-539.

6. Bayles, K. A., & Boone, D. R. (1982). The potential of language tasks for identifying senile dementia. Journal

of Speech and Hearing Disorders, 47, 210-217.

7. Bayles, K.A., & Tomoeda, C.K. (1983). Confrontation naming impairment in dementia. Brain and Language, 19,


8. Bock, K. (1989). Closed-class immanence in sentence production. Cognition, 31, 163-186.

9. Bushell, C.M., & Martin, A. (1997). Automatic semantic priming of nouns and verbs in patients with

Alzheimer disease. Neuropsychologia, 35(8), 1059-1067.

10. Croisile, B. et al. (1996). Comparative study of oral and written picture description in patients with Alzheimer

disease. Brain and Language, 53, 1-19.

11. Fung, T.D., Chertkow, H., Murtha, S., Whatmough, C., Peloquin, L., Whitehead, V., & Templeman, F.D. (2001).

The spectrum of category effects in object and action knowledge in dementia of the Alzheimer

type. Neuropsychology, 15(3), 371-379.

12. Gonnerman, L.M., Andersen, E.S., Devlin, J.T., Kempler, D., & Seidenberg, M.S. (1997). Double dissociation

of semantic categories in Alzheimer disease. Brain and Language, 57, 254-279.

13. Grossman, M., et al. (2003). Neural basis for semantic memory difficulty in Alzheimer disease: an fMRI

study. Brain, 126, 292-311.

14. Hier, D., Hagenlocker, K., & Shindler, A. (1985). Language disintegration in dementia: Effects of etiology

and severity. Brain and Language, 25, 117-133.

15. Hodges, J., & Patterson, K. (1994). Is semantic memory consistently impaired early in the course of

Alzheimer disease? Neuroanatomical and diagnostic implications. Neuropsychologia, 33(4), 441-459.

16. Kempler, D., Curtiss, S., & Jackson, C. (1987). Syntactic preservation in Alzheimer disease. Journal of Speech

and Hearing Research, 30, 343-350.

17. Martin, A., Haxby, J.V., Lalonde, F.M., Wiggs, C.L., & Ungerleider, L.G. (1995). Discrete cortical regions

associated with knowledge of colour and knowledge of action. Science, 270, 102-105.

18. Martin, A., Wiggs, C.L., Ungerleider, L.G., & Haxby, J.V. (1996). Neural correlates of category-specific

knowledge. Nature, 379, 649-652

19. Martin, A., & Chao, L. (2001). Semantic memory and the brain: structure and processes. Journal of

Cognitive Neuroscience, 11, 194-201.

20. Memory. (2003). Stanford Encyclopedia of Philosophy. Retrieved on April 12, 2007 from, http://plato.stanford.


21. Price, T. R. & Berndt, R. S., Mitchum, C. C (2001). Short-term memory and sentence comprehension:

An investigation of a patient with crossed aphasia. Brain, 114, 263-280.

22. Pulvermuller, F. (2005). Brain mechanisms linking language and aging. Journal of Cognitive Neuroscience, 6,


23. Robinson, K., Grossman, M., White-Devine, T., D-Esposito, M. (1996). Category-specific difficulty naming with

verbs in Alzheimer disease. Neurology, 47, 178-182.

24. Silveri, M., Daniele, A., Giustolisi, & Gainotti, G. (1991). Dissociation between knowledge of living and

nonliving things in dementia of the Alzheimer type. Neurology, 41, 545-546.

25. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement,

image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning

and Memory, 6, 174-215.

26. Tippet, L., Grossman, M., & Farah, M. (1996). The semantic memory impairment of Alzheimer disease:

category-specific? Cortex, 32, 143-153.

27. Wechsler, D. (1987). Wechsler memory scale revised. New York: Psychological Corporation.

28. Williamson, D., Adair, J., Raymer, A., & Heilman, K. (1998). Object and action naming in Alzheimer disease.

Cortex, 34, 601-610.


Back to the Journal of Undergraduate Research

College of Liberal Arts and Sciences I University Scholars Program I University of Florida I J:' UNIVERSITY of
SUn yThef FUloiIriIdar The Gil2,r N5hin2
� University of Florida, Gainesville, FL 32611; (352) 846-2032.