ArticlePDF Available

Measurement and Correlates of Verbosity in Elderly People


Abstract and Figures

Two studies were conducted to develop measures of verbosity in elderly people and to determine the social and psychological correlates of verbose speech. In the first study, 346 elderly people were classified into three categories of verbosity on the basis of their verbal behavior during an interview and questionnaire session. Personality variables, stress in daily living, and age differentiated extremely verbose individuals from others. In the second study, frequency and extent of off-target speech were rated quantitatively for the verbal behavior of 203 older men, with a second rater independently making the same ratings for 98 of the men. Classification into the three categories of verbosity was made for 179 of the men. Interrater reliability was established at .76 and .70 for the two measures of verbosity. There was significant agreement between the qualitative classification and the quantitative rating assessments of verbosity. In addition to the previously found associations between verbosity and personality and social variables, higher nonverbal intellectual performance scores obtained in the early adult years combined with poorer current nonverbal scores predicted verbosity in late life.and to determine the social and psychological correlates of verbose speech. In the first study, 346 elderly people were classified into three categories of verbosity on the basis of their verbal behavior during an interview and questionnaire session. Personality variables, stress in daily living, and age differentiated extremely verbose individuals from others. In the second study, quantitative ratings of frequency and extent of off-target speech were made for the verbal behavior of 203 older men, with a second rater independently making the same ratings for 98 of the men. Classification into the three categories of verbosity was made for 185 of the men. Rater reliability was established at .76 and .70 for the two measures of verbosity. There was significant agreement between the qualitative classification and the quantitative rating assessments of verbosity. In addition to the previously found associations between verbosity and personality variables, stress and age, greater declines in non-verbal intelligence from early adult levels predicted verbosity in late life. (Author/KS)
Content may be subject to copyright.
Journal of Gerontology:
Vol. 43. No. 2, P27-33Copyright 1988 by The Gerontological Society of America
Measurement and Correlates of Verbosity
in Elderly People
Dolores Gold, David Andres, Tannis Arbuckle, and Alex Schwartzman
Psychology Department and Centre for Research in Human Development, Concordia University.
Two studies were conducted to develop measures of verbosity in elderly people and to determine the social and
psychological correlates of verbose speech. In the first study, 346 elderly people were classified into three categories of
verbosity on the basis of their verbal behavior during an interview and questionnaire session. Personality variables,
stress in daily living, and age differentiated extremely verbose individuals from others. In the second study, frequency
and extent of off-target speech were rated quantitatively for the verbal behavior of 203 older men, with a second rater
independently making the same ratings for 98 of the men. Classification into the three categories of verbosity
for 179 of the men. Interrater reliability was established at .76 and .70 for the two measures of verbosity. There was
significant agreement between the qualitative classification and the quantitative rating assessments of verbosity. In
addition to
previously found associations between verbosity and personality and social variables, higher nonverbal
intellectual performance scores obtained in the early adult years combined with poorer current nonverbal scores
predicted verbosity in late life.
HIS article reports two studies on off-target verbosity, a
dimension of verbal behavior that can be measured and
that appears to be associated differentially with psychologi-
cal functioning in elderly people. We define off-target ver-
bosity as speech that, although perhaps prompted initially by
an external conversational stimulus, quickly becomes a
prolonged series of loosely associated recollections increas-
ingly remote from, relatively unconstrained by, and irrele-
vant to the present external contextual stimuli. The failure to
maintain focus, which is evident in this continuous intrusion
of irrelevant content, serves to distinguish off-target verbos-
ity from speech, that, although prolonged, remains focused
on elaborating a concept or developing a narrative sequence.
Reminiscence has long been reported to be common
among elderly people (Lewis & Butler, 1974). However,
apart from recognizing its utility in maintaining constancy of
identity, the reminiscence literature appears to report few
well supported conclusions. Furthermore, verbalized remi-
niscence that intrudes excessively and inappropriately into
conversation is characteristic only of a minority of elderly
people, and has not been investigated. Verbosity also has
been described as a symptom in psychopathojogy. The
Diagnostic and Statistical Manual of Mental Disorders (3rd
Ed.) (DSM III) discusses circumstantiality, but acknowl-
edges that such speech is not necessarily associated with
pathology and, indeed, could be common in people who
have no mental disorders (APA, 1980).
The results of some studies of speech and communication
in healthy and dementing elderly people appear relevant to
off-target verbosity. Obler (1980) compared healthy groups
of 50- and 60-year-olds and found that the narrative dis-
course of the older group was more elaborate, used more
words, produced more embedded sentences but fewer
themes, and appeared to be more diffuse, resulting in greater
amounts of speech. The speech of the healthy older group
fell midway between the speech of Parkinsonian patients and
the healthy younger group on these measures. The hy-
pothesis that age is associated with increased wordiness is
supported also by Botwinick and Storandt (1974), who
compared the responses of old and young adults to a vo-
cabulary test. The two age groups did not differ in the
number of "good'' definitions produced, but the older adults
tended to give more multiword definitions and fewer one-
word synonyms.
Hutchinson and Jensen (1980) found that the speech of
dementia patients was characterized by the abrupt intro-
duction of unrelated themes disregarding the cooperative
principle (Grice, 1975), which directs appropriate interac-
tion so that conversation can proceed in a mutually agreed
upon direction. They argued that such verbal behavior repre-
sented egocentrism similar to the egocentrism of early child-
hood. In both cases, the egocentrism was attributed to
limited cognitive capacities; however, the egocentrism of
the demented elderly group was attributed specifically to
decreases in previously adequate cortical capacity. The ex-
tended, irrelevant speech in off-target verbosity also violates
the cooperative principle, and, thus, might also reflect
Studies of verbal memory in elderly patients also have
reported relevant findings. Increased vulnerability to inter-
ference in verbal memory has been reported for healthy
elderly samples (e.g., Moscovitch & Winocur,
cur & Moscovitch, 1983) as well as for samples in the early
stages of Alzheimer's disease (Fuld, Katzman, Davies, &
Terry, 1981). Spilich (1983) found that recall of prose
passages by memory-impaired elderly adults differed quali-
tatively and quantitatively from that of normal elderly and
young adults, being relatively insensitive to the thematic
structure of the passage. Spilich explained the differences as
being due to the memory-impaired elderly adults not being
able to inhibit the entry of unwanted information from long-
term memory into working memory and, as a result, becom-
ing "unstuck from an appropriate contextual framework"
at Concordia University on January 8, 2014 from
Some of these data come from clinical samples and might
not be generalizable to community-based, apparently
healthy elderly people. However, the clinical findings are
suggestive in that the off-target verbosity we have observed
in community-dwelling elderly people shows a similar pat-
tern of intrusion of memories, which are verbalized regard-
less of the conversational context.
Several exploratory hypotheses were developed in rela-
tion to off-target verbosity. It was hypothesized that verbos-
ity and adjustment could be associated either positively or
negatively (Lowenthal, Thurner, & Chiriboga, 1975; Mc-
Mahon & Rudick, 1967) because verbalized reminiscence
could either serve a therapeutic function or, alternatively,
indicate unsatisfactory resolution of earlier developmental
issues. It also was hypothesized that verbosity might reflect
egocentricity caused by isolation (Norris & Rubin, 1984),
or that it might be used to cope with loneliness associated
with aging (Epstein, 1980). The latter explanation was given
by every individual working with elderly people in a variety
of settings with whom this phenomenon was discussed.
These people said they believe that isolated and lonely older
individuals become more demanding in their social interac-
tions in an attempt to compensate for their reduced social
contacts. Thus, it was hypothesized that verbosity would be
greater in people who had less extensive social networks.
The final hypothesis predicted that verbosity would be re-
lated to declining cognitive function, which reduces the
individuals' ability to reorganize information into a logical
and coherent message.
Participants. The sample was made up of 346 Cana-
dian adults, with equal proportions of women and men,
Francophones and Anglophones, and individuals with both
middle- and working-class backgrounds (Blishen & Mc-
Roberts, 1976). The participants were recruited through
community agencies for a study of personality and well-
being. Personality, adjustment, and social functioning mea-
sures were available for all 346 volunteers, whereas cogni-
tive measures were available for a subsample of 285
volunteers. The volunteers ranged in age from 65 to
a mean age of 72.6 and a mean educational level of 11.0
The participants lived independently in their own
homes in Montreal and were tested there.
Procedure. The participants were visited twice for a
total of approximately 3 hr by one of five female interview-
ers who had been trained to interview and administer the
tests in a standardized fashion in either English or French.
The interviewers had been trained to adhere as strictly as
possible to the interview and questionnaire items. If
ipant became verbose, the interviewers were instructed not
to cut off the individual's speech abruptly but to return
persistently to the scheduled order of items as soon as
possible. In the first visit, a life interview was conducted,
and the participants were given a battery of psychological
tests to complete. A series of measures of verbal memory
were completed by the participants in the second visit.
At the end of the second interview, on the basis of their
interaction with the volunteers and without knowing the test
results, the interviewers independently classified the partici-
pants into one of three categories of verbosity: extreme talker
(very verbose, speech consists of sequences of loosely asso-
ciated topics, much extraneous information), controlled
talker (chatty, talks a great deal but sticks more or less to the
point even though much extra information is given), and
nontalker (only answers questions asked and volunteers little
extraneous information). Participants were classified as ex-
treme talkers only when their total session time was ap-
proximately double the average session time and when their
responses were a series of loosely associated topics bearing
little relation to the original question. Interviewers had the
option of not categorizing a participant if they were
One year later, follow-up telephone calls were made to
200 of the volunteers by a different interviewer to discuss
their participation in future research. This interviewer inde-
pendently classified the participants into the same three
categories based on global impression of verbosity.
Measures. The participants were given, in English or
French, standardized measures that have been used exten-
sively with elderly samples. The Eysenck Personality Inven-
tory (EPI) (Eysenck & Eysenck, 1968) was used to measure
extraversion, neuroticism, and lie scores, or tendency to
respond in a socially desirable way. The Memorial Univer-
sity of Newfoundland Scale of Happiness (MUNSH)
(Kozma & Stones, 1983) was used to assess psychological
well-being. The Desire and Belief Scale (Reid & Ziegler,
1980) was used to assess the extent to which reinforcers were
desired by the individual and the perceived locus of control
for these reinforcers. A checklist was used to assess the
activities in which the participants regularly engaged,
weighting each activity by frequency and the social effort
involved (Arbuckle, Gold, & Andres, 1986). A short form
of the Holmes and Rahe Social Readjustment Scale (1967),
retaining only items appropriate for older samples, was used
to measure the number of stressful events experienced in the
past year. The individuals also rated the adequacy of differ-
ent aspects of their lives, such as financial situation and
health, for everyday functioning (Schonfield & Hooper,
Three different tasks provided measures of memory
and verbal comprehension: forward-digit span; memory and
comprehension of prose; and organization and free recall of a
list of categorizable words. Recall scores on the tasks were
summed to provide a memory index, and the Modified Ratio
of Repetition scores for clustering measured organization
(Bower, Lesgold, & Tienam, 1969). Relations among the
measures of cognitive, social, and personality functioning
for this sample have been described elsewhere (Arbuckle et
Of the 346 participants, 74 (21.4%) were classified as
extreme talkers, and 156 (45%) were classified as non-
talkers. The remaining 116 (33.5%) participants were de-
scribed as controlled talkers. Chi-square analysis indicated
at Concordia University on January 8, 2014 from
that marital status, gender, and ethnicity did not differ
among the groups. A multivariate analysis of variance
(ANOVA) followed by univariate ANOVAs revealed a sig-
nificant multivariate difference between the groups,
F(23,198) = 1.94, p < .01, with extreme talkers having
significantly lower scores on the desire for reinforcers scale,
< .05, w2 = .022; lower self-ratings on
adequacy of finances, F(\,222) = 4.23, p < .05, w2 =
and being older, F(l,222) = 6.47, p < .05, co2 =
than nontalkers. Other differences showed that ex-
treme talkers were more extraverted, F(l,222) = 3.38, p <
co2 = .01; had lower social desirability set scores,
F(l,222) = 3.38, p < .10, co2 = .01; and had more
difficulty with transportation, F(
= 3A2,p< .10, oo2
= .009 than nontalkers. To clarify the finding of significant
differences on the desire for reinforcers scale, a post hoc
factor analysis was performed on this scale. This analysis
produced a first factor, accounting for 70% of the variance,
largely composed of items assessing the individuals' desire
for independence in their activities. Extreme talkers scored
significantly lower on a scale composed of items that loaded
on this factor, F(
= 6.78,/? < .05, w2 = .025, but did
not differ significantly from nontalkers on the other items.
Percentage agreement among categories was used to de-
termine the extent of agreement between the interview and
telephone classifications of the participants into the verbos-
ity categories, lnterrater agreement on classification was
To summarize, the verbose older people did not differ
with regard to social class, gender, marital status, ethnicity,
adjustment, social activities, happiness, contact with family,
memory functioning, and self-perceived health from those
people whose speech was appropriate for social interaction.
Therefore, the results do not support the hypotheses that
excessive speech is found in older people who are more
isolated, who differ in adjustment, or who have poorer
memories. Age, stress, and personality variables were the
main differentiators of verbosity.
Agreement between raters in classifying participants on
verbosity was low. This might be due to changes in verbal
behavior that occurred during the interval between assess-
ments, to differences in verbal behavior over the telephone
as opposed to a person-to-person situation, and to different
interview topics. However, it is likely that the subjectivity of
the global classification imposed inherent constraints on its
reliability. The second study was designed to develop a more
reliable quantitative method and to determine the extent of
agreement between the two methods. The study also was
designed to determine if the age, stress, and personality
factors that had emerged as correlates of verbosity in the first
study would do so again. Finally, additional measures of
functioning, including archival data, were included to fur-
ther understanding of the phenomenon.
Participants. —The participants in this study were World
War II Canadian Army veterans who had volunteered to
participate in a study of psychological functioning. With the
cooperation of Canada Archives, we sent letters to veterans
in Quebec and Ontario who had been in contact with this
agency concerning their army service, and in the letters we
briefly described the study and asked them to participate.
Cooperating veterans associations in Quebec and Ontario
also sent the same letter to their membership. Finally, some
veterans volunteered in response to articles describing the
study that were published in local community newspapers
and in the Canadian Legion magazine. The participants were
required to take part in a life-history interview, to complete
measures of psychological and social functioning in either
English or French, and to allow access to their army files.
The volunteers were paid $20 for participating. Data relating
to verbosity were collected for 203 veterans, 137
Anglophones and 66 Francophones, and are reported here.
All participants were males. The sample ranged in age
from 60 to 81, with a mean age of 65.30 years and a mean
educational level of 10.10 years. Blishen socioeconomic
occupational prestige scores for the sample averaged 46.52,
indicating an employment background that was predomi-
nantly upper working and lower middle class (Blishen &
McRoberts, 1976).
Measures. The participants were given standardized
measures, including some used in the previous study. These
were the EPI (Eysenck & Eysenck, 1968), the activities
checklist (Arbuckle et al., 1986), the MUNSH (Kozma &
Stones, 1983), the Desire and Belief Locus of Control Scale
(Reid & Ziegler, 1980), and self-ratings of daily functioning
(Schonfield & Hooper, 1973). In addition, the Life Experi-
ences Survey (LES) (Sarason, Johnson, & Siegel, 1982) was
used to measure life stress, which was defined as the impact
of the total number of life changes occurring in the past year
of the participant's life. Scores reflected the total number of
events and the rating of the perceived positive and negative
impact of each event. The Social Support Questionnaire
(SSQ) (Sarason, Levine, Basham, & Sarason, 1983) was
used to assess the extent of and perceived satisfaction with
the support networks available to the individuals. The partic-
ipants listed all persons who provided them with support in
different situations and their satisfaction with the support
received. Health was assessed by the Seriousness of Illness
Inventory (Wyler, Masuda, & Holmes, 1971); the total score
reflects the number and seriousness of disease syndromes the
individual had, with seriousness determined by health pro-
fessional ratings. Forward-digit span was used as a measure
of immediate memory.
The participants also completed the Revised Examination
M Test of Intelligence, which they had taken for the first
time approximately 40 years before upon their induction into
the army. The M test was developed in French and English
versions by the Canadian Army to help classify recruits. The
scale includes nonverbal subtests (picture completion, pic-
ture anomalies, and paper formboard) and verbal subtests
(arithmetic, vocabulary, and verbal analogies). Validation
studies conducted by the Personnel Service of the Canadian
Army reported a correlation coefficient of .80 between the M
test and the American Army Alpha Test (W. R. N. Blair,
personal communication, 1959). We have found total test-
at Concordia University on January 8, 2014 from
retest reliability for this sample to be .78 over 40 years
(Schwartzman, Gold, Andres, Arbuckle, & Chaikelson,
Procedure. Participants were given an interview, re-
corded on audiotape, that gathered demographic and per-
sonal data, focusing on work and family history. Verbose
speech was defined operationally as speech that was off
target when replying to a stimulus in the interview, that is,
speech that conveyed information irrelevant to the question.
This definition excludes speech that remains focused on the
original stimulus and that develops a relevant response. Two
scores were derived for off-target speech. Item verbosity
consisted of the number of interview items on which the
participant had been rated as going off target. In addition,
each time the individual went off target, the interviewer
rated the extent to which the person strayed from the topic on
a 5-point Likert scale. These ratings were summed to calcu-
late the second score, extent of verbosity. A second pair of
scores for verbosity, this time for a noninterview situation,
was obtained by using the same procedures that were used
when the participants completed the LES. At the end of the
session, the interviewer rated the respondent on global
impression of verbosity, according to the same three catego-
ries derived in the first study: nontalker, controlled talker,
and extreme talker. The interviewers also had the option of
giving a "don't know" response. Finally, the interviewers
recorded the duration of the entire session. An independent
rater subsequently scored 48% of the tapes on the two
quantitative measures, item verbosity, and extent verbosity.
Interrater reliability coefficients were calculated between
the verbosity ratings of the two independent raters for 98
interviews. There was significant agreement between the
with coefficients of .76 for number of items and .70
for extent of verbosity being obtained (p < .001). Item
verbosity and extent verbosity were highly correlated in both
the interview (r = .89,/? < .001) and the LES (r = .87,/? <
.001) situations for the total sample of 203 men. As can be
seen in Table 1, the verbosity item and extent measures were
significantly correlated across the interview and LES situa-
Since the correlations for the item and extent measures
in the two situations were significant but moderate, they
were combined to form one set of verbosity measures, total
number of items and total extent of verbosity. These two
scores correlated significantly (r = .90, p < .001).
Table 1. Correlation Coefficients for Verbosity Measures
Interview verbosity
LES verbosityNumber of
of verbosity
Number of items
Extent of verbosity
The interviewers classified 179 participants into one of the
three categories of nontalker, controlled talker, and extreme
talker. As can be seen in Table 2, 18% of the participants
were categorized as extreme talkers. One-way ANOVAs
were conducted comparing mean scores on the six items and
extent verbosity measures across the three classifications.
The three groups differed significantly on all of the verbosity
measures, F(2,178) = 8.73, p < .001, a>2 = .11, and,
F(2,178) = 11.28,/?< .001, (o2 = .12 for item and extent
verbosity in the interview situation, F(2,178) = 6.00, p <
a)2 = .06;F(2,167) =
< .001, w2 = .09, for
item and extent verbosity in the LES situation; F(2,178) =
p < .001, a)2 = .11, and, F(2,178) = 14.50, p <
oo2 = .15, for the total item and total extent verbosity
measures, respectively. Duration of time of the interview
and test session also differed significantly for the three
classifications, F(2,177) = 41.29,/? < .001, co2 = .33. As
Table 2 indicates, Scheffe tests revealed that the nontalkers
were significantly less verbose on all six measures and
required less time to complete the session (p < .05). Ex-
treme talkers had a significantly higher extent of verbosity
ratings in the interview and in their total extent of verbosity
ratings than did the controlled talkers (p < .05). Extreme
talkers also required more time for the session than did those
in the other two groups (p < .05).
Two multiple-regression analyses were performed to de-
termine the demographic and psychological variables that
significantly predicted total item and total extent verbosity.
socioeconomic status scores; marital status; language;
desire and expectancy scores; LES scores; activities; extra-
version, neuroticism, and lie EPI scores; digit span, and
availability of and satisfaction with social support scores
were entered into the regression. In addition, wartime and
current total scores for nonverbal and verbal performance on
the M test also were entered into the regression. To assess
Table 2. Analysis of Variance (ANOVA) Results,
by Verbosity Classification
Interview item
Interview extent
LES item
LES extent
Total item
Total extent
Duration of
(w = 64)
6.03 (5.65).
.46 (.91).
.57 (1.12).
6.49 (6.08).
188.93 (33.62).
(n = 83)
9.34 (7.07)b
17.24 (15.46)b
1.44 (2.36)b
2.08 (3.51)b
10.78 (8.54)b
19.32 (16.31)b
221.06 (37.29)b
(n = 32)
11.62 (7.10)b
1.58 (1.65)b
3.20 (3.75)b
13.56 (7.75)b
28.29 (19.37)c
270.03 (59.14)c
Note. N = 203. LES = Life Experiences Survey.
*p< .001.
Note. Values are means plus or minus standard deviation. LES = Life
Experiences Survey.
"Duration measured in minutes.
Values with different subscripts differ significantly atp < .05.
at Concordia University on January 8, 2014 from
the predictive power of the most important variables in the
study, we entered the subset of significant predictors of
verbosity that emerged from the initial regressions into two
subsequent reduced regressions.
A total of 12.8% of the variance in total item verbosity
was accounted for by six variables in the study. Table
the standardized Betas for the predictors, the zero-order
correlation coefficients, and the
coefficients indicating the
particular effect of each variable. Wartime and current
nonverbal performance scores, EPI lie and extraversion
scores, activities, and stress predicted total item verbosity.
Men who scored lower on current measures of nonverbal
ability but who had scored higher on the nonverbal section of
the M test 40 years ago, were more likely to go off
addition, men who were more socially active, who demon-
strated less concern for socially desirable self-presentation,
who were more extraverted, and who had experienced more
stress were verbose more frequently. The results are similar
for the second measure, total extent verbosity, with 9.8% of
the variance being accounted for by five of the study vari-
Current and wartime nonverbal scores on the M test,
EPI lie scores, and activities were significant predictors,
whereas extraversion scores predicted verbosity at the .10
level. Men who had scored higher 40 years ago but who
scored lower on current measures of nonverbal intellectual
ability, who scored lower on the social desirability scale,
who were more socially active, and who were more extra-
verted displayed a greater degree of verbosity.
Wartime and current total scores for nonverbal intellectual
which correlated at .58, were examined further to
clarify their role in predicting verbosity. Two-way
ANOVAs comparing veterans who were above or below the
median on wartime and current nonverbal total scores were
calculated on total item and total extent verbosity scores.
Both the interaction between wartime and current nonverbal
total scores and the main effect of wartime nonverbal total
Table 3. Results of Multiple-Regression Analyses
of Verbosity Scores
Total item verbosity
Current M test nonverbal scores
Wartime M test nonverbal scores
Extraversion scores
EPI lie scores
= .128
Total extent verbosity
Current M test nonverbal scores
Wartime M test nonverbal scores
EPI lie scores
Extraversion scores
Total R2 = .098
•Stress was scored so that lower scores indicate more negative impact.
*p< .10, **p< .05, ***p< .01.
scores were significant predictors for total item verbosity,
F(l,202) = 4.71, p < .05, F(l,202) = 5.02, p < .05,
respectively. A Scheffe test of individual comparisons re-
vealed that only those veterans who had scored above the
median for wartime and who were now below the median on
current nonverbal scores were higher on total item verbosity,
whereas the other three groups did not differ significantly
among themselves (p < .05). Similar results were obtained
for the analysis of total extent verbosity, with the interaction
between wartime and current nonverbal scores significantly
predicting verbosity, F(l,202) = 5.34, p < .05. Those
veterans who had scored above the median and who cur-
rently were below the median on nonverbal total scores were
significantly higher on total extent verbosity than were those
in the other three groups, who did not differ among them-
selves (p < .05).
The results of the two studies have several noteworthy
aspects, both with regard to the measurement of and the
variables associated with verbosity. The substantial agree-
ment between the two quantitative measures, both within
and across the interview and questionnaire situations, indi-
cates that verbose individuals tend to be so consistently. It is
not just a few topics that stimulate off-target speech; rather,
such speech seems to stem from a generalized set so that it
ends up reflecting inner preoccupation rather than external
stimuli. Furthermore, such preoccupation becomes verbal-
ized not only in the interview session where it might be
stimulated by open-ended questions, but also in the ques-
tionnaire session where it is less likely to be elicited by the
procedure. The results also clearly indicate that verbosity
can be assessed quantitatively with a high degree of agree-
ment by different raters.
It is evident from the results of the two studies reported
here that, to a significant extent, both the qualitative, global
method and the more molecular, quantitative approach sam-
ple the same process. Not only was the group classification
predictive of significant differences in quantitative scores,
but a consistent pattern of variables was associated with
verbosity as measured by the two methods.
The hypothesis emerging from the first study predicted
that verbosity would be associated with older age, lesser
desire for independent functioning, lesser concern for pre-
senting oneself in a socially desirable way, greater extraver-
sion, and experiencing more stress. The results of
study partially replicated and extended those findings with a
somewhat younger sample. In the latter study, lesser con-
cern with making a socially desirable impression on others,
extraversion, greater social activity, stress, and the combi-
nation of higher young adult levels of nonverbal intellectual
performance with poorer current scores predicted higher
frequency of verbosity. With the exception of stress, these
variables also predicted greater extent of verbosity. How-
ever, it is clear that the majority of variance in off-target
verbosity has not yet been explained.
Age was eliminated as a significant predictor of verbosity
in the second study only when scores of nonverbal ability
were entered into the regression. Thus, verbosity appears to
be related not to age but to decreases in nonverbal intel-
at Concordia University on January 8, 2014 from
lectual ability. The nonverbal subtests, which sample the
ability to monitor and understand the environment and to
perform abstract reasoning, appear conceptually similar to
measures of fluid intelligence, which declines more rapidly
with age than do other aspects of intelligence (Horn, 1982).
The finding that verbosity is predicted by declining per-
formance from an initially higher level of nonverbal intel-
ligence in young adulthood is interesting. Inasmuch as
wartime verbal and nonverbal total scores were significantly
correlated (r = .59), verbose men might have been some-
what more proficient verbally and nonverbally 40 years ago.
Nonverbal scores decreased significantly over the 40-year
period, whereas verbal intelligence scores did not
(Schwartzman et al., 1987). Furthermore, neither wartime
nor current verbal scores predicted the level of verbosity.
This pattern of results suggests that failure to maintain focus
in speech could reflect the particular combination of a sig-
nificant loss of nonverbal intellectual ability and intact, well-
developed, and practiced verbal skills. This hypothesis will
be tested in future research.
Variables assessing difficulties in functioning predicted
verbosity in the two studies. In the first study, specific
indicators of stress were linked to excessive speech. In the
second study, a general life events measure revealed that
verbose individuals had undergone greater negative rami-
fications of changes in their lives during the past year.
Personality variables also emerged consistently as signifi-
cant predictors of verbosity in both studies. For both sam-
a lesser tendency to present oneself in a socially
desirable fashion and higher extraversion predicted verbos-
ity. Lack of concern for social presentation of self can be
interpreted as indicating egocentrism. Social activities also
emerged as a significant predictor of verbosity in the second
study, whereas desire for independent functioning was not a
salient variable for the younger sample. It should be empha-
sized that we found no support for the hypothesis consist-
ently advanced by those who work with elderly people: that
talkative old people are lonely. In fact, the more verbose
individuals in our study reported being more socially active.
Our research to this point establishes that off-target ver-
bosity can be reliably measured and that it has consistent
associations; with psychological functioning. It becomes
manifest in .older people who are extraverted, socially ac-
not concerned with others' impressions of
going more stress and experiencing declining performance
in nonverbal cognitive functioning. The function that ver-
bosity serves for these individuals is still unspecified. Blythe
(1981) postulated that incessant autobiographical talk is a
way of maintaining identity by elderly people at a time when
many significant signposts of their lives are being obliter-
ated. Recounting of past experience could serve as a defen-
sive reaction, reassuring the older person that his or her
identity remains the same (Lieberman & Tobin, 1983).
It is not clear if the effects of the personality and social
variables associated with verbosity are independent of or
interact with biological aging. It is possible that older indi-
viduals who are extraverted, socially active, and less con-
cerned with others' impressions of them have been inclined
to be more talkative throughout their lives. This hypothesis
is supported by research indicating that extraversion is a
stable personality trait across the adult life span (McCrae &
Costa, 1984). The verbal behavior of such people might
become excessive only as they grow old and encounter the
stresses associated with aging. Thus, verbosity could be a
relatively benign behavioral variant, occurring in elderly
people who have certain personality and social characteris-
tics and serving to maintain constancy of identity. Alterna-
tively, the pattern of poorer cognitive functioning associated
with verbosity could be indicative of central nervous system
impairment. This hypothesis will be examined in future
This research was supported by grants from the Social Sciences and
Humanities Research Council of Canada, the Quebec Council for Social
Affairs, and the Natural Sciences and Engineering Council of Canada. Parts
of these data were presented at meetings of the Canadian Psychological
Association, June 1985, and the Canadian Gerontology Association, No-
vember 1986.
We wish to express our appreciation to Claude Senneville, Anastassos
Stalikes, and Lina Bortolussi for their work on the project and to the
agencies and volunteers for their cooperation with the study.
Address correspondence to Dr. Dolores Gold, Centre for Research in
Human Development, Concordia University, 1455 de Maisonneuve Boule-
vard West, Montreal, Quebec, Canada, H3G 1M8.
American Psychiatric Association (1980). Diagnostic and statistical man-
ual of mental disorders (3rd ed.). Washington, DC: Author.
Arbuckle, T. Y., Gold, D., & Andres, D. (1986). Cognitive functioning of
older people in relation to social and personality variables. Psychology
and Aging, 1, 55-62.
Blishen, B. R., & McRoberts, H. A. (1976). A revised socioeconomic
index for occupations. Canadian Revue of Sociology and Anthropology,
Blythe, R. (1981). The view in winter: Reflections on old age. London:
Botwinick, J., & Storandt, M. (1974). Vocabulary ability in later life.
Journal of Genetic Psychology, 125, 303-308.
Bower, G. H., Lesgold, A. M., &Tienam, D. (1969). Grouping operations
in free recall. Journal of
Learning and Verbal Behavior, 8, 481-
Epstein, L. (1980). Issues in communication in geropsychiatry. In L. K.
Obler & M. L. Albert (Eds.), Language and communication in the
elderly. New York: Lexington Books.
Eysenck, H. J., & Eysenck, S. B. G. (1968). Manual: EysenckPersonality
San Diego, CA: Educational and Industrial Testing Service.
Fuld, P.,Katzman, R., Davies, P., & Terry, R. D. (1981). Intrusions as a
sign of Alzheimer dementia: Chemical and pathological verification.
Annals of Neurology, 11. 155-159.
Grice, H. P. (1975). Logic and conversation. In P. Cole & J. Morgan
Syntax and semantics: Speech acts (Vol. 3). New York: Aca-
demic Press.
Holmes, T. H., & Rahe, R. H. (1967). The social readjustment rating scale.
Journal of Psychosomatic Research, 11, 213-218.
Horn, J. L. (1982). The aging of human abilities. In B. Wolman & G.
Strieker (Eds.), Handbook of developmental psychology. Englewood
NJ: Prentice-Hall.
Hutchinson, J. M., & Jensen, M. (1980). A pragmatic evaluation of
discourse communication in normal and senile elderly in a nursing
home. In L. K. Obler & M. L. Albert (Eds.), Language and communi-
cation in the elderly. New York: Lexington Books.
Kozma, A., & Stones, M. J. (1983). Predictors of happiness. Journal of
Gerontology, 38, 626-628.
Lewis, M. I., & Butler, R. N. (1974). Life review therapy: Putting
memories to work in individual and group psychotherapy. Geriatrics,
29, 165-173.
at Concordia University on January 8, 2014 from
M. A., &
S. S.
Stress, coping and survival. New York: Basic Books.
Lowenthal, M. F., Thurner, M., Chiriboga,
(1975). Four stages of life:
A comparative study
women and
facing transitions.
CA: Jossey-Bass.
& Costa, P.
Jr. (1984). Emerging lives: Enduring disposi-
tions. Boston, MA: Little Brown & Company.
A. W., &
P. J.
(1967). Reminiscing
in the
adaptational response. In S. Levin & R. Kahana(Eds.), Psychodynamic
studies and aging: Creativity, reminiscing
dying. New York: Interna-
tional Universities Press.
Moscovitch, M., & Winocur,
(1983). Contextual cues and release from
proactive inhibition
and old
people. Canadian Journal
J. E.,
& Rubin,
(1984). Peer interaction and communication.
& O. G.
(Eds.), Lifespan development
behavior (Vol. 6). Orlando, FL: Academic Press.
L. K.
(1980). Narrative discourse style in the elderly.
M. L.
Albert (Eds.), Language
in the
New York: Lexington Books.
D. W.,
& Ziegler,
(1980). Validity and stability
a new desired
control measure pertaining
psychological adjustment
the elderly.
Journal of Gerontology,
I. G.,
J. H., &
J. M.
(1982). Assessing
life changes.
In 1. G.
& C. D.
Spielberger (Eds.),
Stress and anxiety (Vol. 6). New York: Wiley.
I. G.,
H. M.,
R. B., &
B. R.
Assessing social support: The social support questionnaire. Journal of
Personality and Social Psychology,
D., &
(1973). Future commitments and successful
aging: Special groups. Journal of Gerontology,
A. E.,
T., &
J. (1987). Stability
40-year follow-up. Canadian
Journal of Psychology, 42(2), 244-256.
G. J.
(1983). Life-span components
text processing: Structural
and procedural differences. Journal
Verbal Learning
Behavior, 231-244.
G., &
(1983). Paired-associate learning
interference and context effects. Journal of Gerontology,
Wyler, A.
Masuda, M., & Holmes, T. H. (1971). The magnitude
illness. Psychosomatic Medicine,
Received May 21, 1987
Accepted October 20, 1987
The Gerontological Society of America is seeking nominations for the position of Editor-in-Chief of its
bimonthly publication, The
The term of office is four years, commencing January 1, 1989.
Applicants should have a substantial publishing record, preferably in scholarly journals, editorial-related
experience, and knowledge of gerontology and the Society. This is an unpaid position. A curriculum vita and
any supporting materials should be sent to: Dr. Edward Masoro, Chairman of the Publications Committee,
c/o GSA Central Office. Deadline for receipt of applications is April
at Concordia University on January 8, 2014 from
This research investigates how elderly patients in Libya interact and communicate institutionally with younger physicians. In communities such as that in Sebha, Libya (Bedouin, Arab and Muslim) elderly people have important social status and power. Naturally occurring conversations between three elderly patients and their younger physicians were tape-recorded separately. The subjects were also interviewed. Conversation Analysis combined with social constructionist theories was employed as a method of data analysis. The analysis of the interviews shows that elderly patients and their younger physicians (subjects) perceive each other socially rather than institutionally. Younger physicians insist that they cannot practice their institutional concessions when the patients are elderly. They cannot produce interruption, overlap or verbal refusal. In comparison, elderly subjects suggest that they perceive their younger physicians like their younger sons. The analysis of the naturally occurring conversations reveals that elderly patients and younger physicians call each other as social actors (son/hajj or uncle) rather than institutional ones (doctor/sir). Elderly patients and younger physicians could not raise sexual topics or directly name sexual organs (e.g. rectum) related to the elderly patient's case. Moreover, elderly patients address their younger physicians with proverbs and religious expressions when describing their cases. The conversations’ exchange of turns was not thoroughly task-oriented. Rather, elderly patients address their younger physicians with social conversations (e.g. where are you from? Who is your cousin? Where do you live?). Elderly patient interactants manipulated the conversations’ time-span. Furthermore, they employed the least preferable repair strategy. This research concludes that elderly patients in Libya interact and communicate with their younger physicians in accordance to their social construction which shows prevalence over the institutional hierarchy of the interactants.
Many of the broad models of cognitive aging are simply too general to account for the diverse range of data. General slowing cannot readily account for processing asymmetries in lexical processing or for differences in outcomes within single tasks. An understanding of the effects of normal aging on language processing requires attention to a complex interaction of processes, from low-level sensory deficits that can affect high-level discourse processes to vice versa. The current models of cognitive aging have typically been applied to only a subset of the phenomena we have discussed at word, sentence, and discourse levels of language comprehension and production. Moreover, the influence of lower level sensory deficits, which can have substantial effects on working memory and higher level comprehension, has generally not played a large role in theory development.
Full-text available
The present research investigated younger and older adults' communicative goals and their effects on off-topic speech for autobiographical narratives. Participants indicated their communicative goals by rating preferences among paired goals, for example, focus-fascinating, one of which was designated as an expressive goal, appropriate for producing elaborative speech, and one of which was an objective goal, suited to producing concise speech. The participants then told stories about episodic and procedural topics, which were rated by groups of younger and older listeners. Age differences emerged in communicative goals, where younger adults clearly favored expressive goals for episodic topics and objective goals for procedural topics. In contrast, older adults' goals were more diverse, consisting of a mixture of expressive and objective goals for both topic types, without a clear preference. Younger adults' goals predicted ratings of off-topic speech assessed by listeners: Younger and older adults were perceived as equivalently focused, coherent, and clear for episodic topics, but older adults were perceived as less focused, less clear, and more talkative than younger adults on procedural topics. These results suggest that age-related changes in off-topic speech emerge as a result of younger adults selecting goals designed to produce more succinct stories.
Full-text available
In two experiments, the effects of taxing selective attention processes on the efficiency of working memory processes were considered in relation to normal aging. In both experiments, the presence of task-irrelevant information disrupted the efficiency of working memory processes, and the effect was generally greater for older than for younger adults. The presence of distracting information increased the frequency of intrusion errors in both younger and older adults and of memory-based errors in older adults. These findings suggest that distraction disrupts both the ability to maintain a coherent stream of goal-directed thought and action in younger and older adults and the encoding and retention of relevant information in older adults.
Several recent studies of aging and cognition have attributed decreases in the efficiency of working memory processes to possible declines in attentional control, the mechanism(s) by which the brain attempts to limit its processing to that of task-relevant information. Here we used fMRI measures of neural activity during performance of the color-word Stroop task to compare the neural substrates of attentional control in younger (ages: 21-27 years old) and older participants (ages: 60-75 years old) during conditions of both increased competition (incongruent and congruent neutral) and increased conflict (incongruent and congruent neutral). We found evidence of age-related decreases in the responsiveness of structures thought to support attentional control (e.g., dorsolateral prefrontal and parietal cortices), suggesting possible impairments in the implementation of attentional control in older participants. Consistent with this notion, older participants exhibited more extensive activation of ventral visual processing regions (i.e., temporal cortex) and anterior inferior prefrontal cortices, reflecting a decreased ability to inhibit the processing of task-irrelevant information. Also, the anterior cingulate cortex, a region involved in evaluatory processes at the level of response (e.g., detecting potential for error), showed age-related increases in its sensitivity to the presence of competing color information. These findings are discussed in terms of newly emerging models of attentional control in the human brain.
Full-text available
Age differences in performance on memory measures and in subjective ratings of memory adequacy were examined in the context of 12 social, personality, adjustment, and lifestyle measures. Participants were 285 men and women, aged 65 to 93, of middle- and working-class backgrounds. A series of multivariate and univariate analyses revealed that a large proportion of the age differences and virtually all of the social-class differences on memory measures could be accounted for by contextual variables, with education, intellectual activity, extroversion, neuroticism, and lie scores (on the Eysenck Personality Inventory) all accounting for more of the variance in memory performance than did age. Self-rated memory adequacy was not correlated with performance, and although the expected finding of lower ratings by older participants was obtained with the working-class group, the opposite was true for the middle-class group. Implications of these results for understanding age differences in memory are discussed.
Describes the Social Support Questionnaire (SSQ) and 4 empirical studies employing it. The SSQ yields scores for (a) perceived number of social supports and (b) satisfaction with social support that is available. Three studies (N = 1,224 college students) dealt with the SSQ's psychometric properties, its correlations with measures of personality and adjustment, and the relation of the SSQ to positive and negative life changes. The 4th study (40 Ss) was an investigation of the relation between social support and persistence in working on a complex, frustrating task. The research reported suggests that the SSQ is a reliable instrument and that social support is (a) more strongly related to positive than negative life changes, (b) more related in a negative direction to psychological discomfort among women than men, and (c) an asset in enabling a person to persist at a task under frustrating conditions. Clinical implications are discussed. (47 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
Compared a group of young adults (17-20 yrs) to a group of older adults (62-83 yrs) to examine the relationships between qualitative and quantitative vocabulary test scores as they involve the age of S. Results of the study cast doubt as to the validity of the generalization that vocabulary function is maintained throughout adult life. The best possible response (i.e., superior synonym) declined with age. This decline was compensated for by very satisfactory, although less good, performance. This made the overall level of correct response appear similar for the 2 age groups, which may account for the usual report of stability of function in the course of adult aging. The compensation also accounted for very high correlations between qualitative and quantitative scores. (PsycINFO Database Record (c) 2012 APA, all rights reserved)