Research suggests that memories triggered by olfactory
information are different than memories evoked by verbal
and visual information (e.g., Herz & Schooler, 2002; Wil-
lander & Larsson, 2006). For example, it is well established
that verbal cuing of memories generates a reminiscence
bump between the age of 10–30 years at event (e.g., Conway
& Haque, 1999; Rubin, 1982; Rubin & Schulkind, 1997;
Rubin, Rahhal, & Poon, 1998). However, in contrast with a
memory peak in young adulthood, olfactory cued memories
have been localized to the first decade of life indicating that
odor representations are older than verbal and visual ones
(Chu & Downes, 2000; Willander & Larsson, 2006).
Likewise, evidence suggests that the phenomenological
experience of olfactory memory differs from the process-
ing of other sensory information. For example, the feeling
of being brought back in time to the occurrence of the
event is experienced as stronger for odor-cued memories
than memories evoked by words and pictures (Herz, 2004;
Herz & Schooler, 2002; Willander & Larsson, 2006).
Also, some research suggests that odor evoked memories
may be experienced as more emotional than memories
cued by other sensory modalities (e.g., Herz, 1998; Herz
& Cupchik, 1995; Herz, Eliassen, Beland, & Souza, 2004;
Herz & Schooler, 2002), although other work has failed
to replicate these observations (Ehrlichman & Bastone,
1992; Willander & Larsson, 2006). A possible explana-
tion for these discrepant findings may be methodologi-
cal differences across studies such that the experiential
qualities have been assessed in conjunction with the
olfactory stimulation or after the stimulus presentation
(Herz & Schooler, 2002; Willander & Larsson, 2006).
Thus, previous results reporting an advantage of odors
as cues to emotional autobiographical memories may be
a result of participants evaluating the olfactory cue rather
than the retrieved memory. The underlying assumption
for the advantage of odors as cues to emotional memo-
ries is the direct synapsing from the olfactory areas to the
amygdala–hippocampal complex, the neuronal substrate
of emotional memory (Herz et al., 2004; Packard, Cahill,
& McGaugh, 1994; Savic, Gulyàs, Larsson, & Roland,
2000). Because passive smelling of odors may influence
the emotional arousal in individuals (e.g., Lehrner, Mar-
winski, Lehr, Johren, & Deecke, 2005; Savic et al., 2000),
without memory recollection, it is of interest to investigate
the role played by olfactory perception in the recollection
and phenomenological evaluation of retrieved memories
(cf. Chu & Downes, 2002; Herz & Schooler, 2002).
An unexplored question concerns whether verbal and
conceptual processing has an impact on the age distribution
of olfactory-evoked memories. A review of related evidence
suggests that olfactory perception is influenced by verbal
knowledge of an odor. For example, Distel and Hudson
(2001) showed that intensity, pleasantness, and familiar-
ity ratings were higher when odor judgments were done in
conjunction with a verbal label as compared with a no-label
condition. Also, Herz and von Clef (2001) reported higher
positive judgments in pleasantness, intensity, and familiarity
when an identical odor was presented with a positive label
(i.e., Parmesan cheese) as compared with a negative label
(i.e., vomit). In a similar vein, research also indicates that
semantic olfactory knowledge, such as perceived familiarity
and identifiability are positively related to episodic retention
1659 Copyright 2007 Psychonomic Society, Inc.
Olfaction and emotion: The case of
Stockholm University, Stockholm, Sweden
This study investigated (1) the influence of verbal and conceptual processing on the retrieval and phenomeno-
logical evaluation of olfactory evoked memories, and (2) whether the experienced qualities of retrieved informa-
tion are affected by olfactory exposure per se. Seventy-two older adults were randomized into one of three cue
conditions (odor only, name only, or odor name) and asked to relate any autobiographical event for the given
cue. The results indicated that semantic knowledge of an odor’s name significantly affects the age distribution
of memories such that the memory peak in childhood observed for odors only was attenuated. Also, experiential
factors such as pleasantness and feelings of being brought back in time were lower when odors were presented
with their respective names. Olfactory evoked memories were associated with a higher emotional arousal that
could not be accounted for by the perceptual stimulation alone. Taken together, the overall pattern of findings
suggests that retrieval of olfactory evoked information is sensitive to semantic and conceptual processing, and
that odor-evoked representations are more emotional than memories triggered by verbal information.
Memory & Cognition
2007, 35 (7), 1659-1663
J. Willander, email@example.com
ularly replaced to ensure freshness. Words (i.e., the names of the
odors) were printed in Arial (size 100 points, lowercase) and pre-
sented on a 12-in. LCD computer screen.
The retrieved memories were rated on the following six dimen-
sions: (1) How pleasant is the event at this moment? (2) How intense
is the event at this moment? (3) How strong is the feeling of being
brought back in time to the occurrence of the event? (4) How vivid is
the memory of the event? (5) How emotional do you experience the
remembered event at this moment? (6) From what point of view did
you recall the event? For Question 1, a 9-point Likert scale (1 very
unpleasant, 5 neutral, 9 very pleasant) was used. For Questions
2 to 5, a 9-point Likert scale (1 not at all, 9 very much) was used.
Question 6 was assessed with the following three alternatives: (1) field
perspective, (2) observer perspective, (3) neither field or observer per-
spective. The third option was included because some participants
were unable to choose a perspective (i.e., field or observer).
All subjects were tested individually and were randomly assigned
to one of the three cue conditions: name only (n 24; 16 women,
8 men), odor only (n 24; 17 women, 7 men), or odor name (n
24; 17 women, 7 men). Participants in the odor-only condition were
presented with 20 odors and asked to relate any autobiographical
event for the given cue. In instances when a memory was retrieved,
the participant was asked to write down a short description of the
event, and to rate its experiential attributes (see above). The same
instructions and rating procedure were applied in the name-only
condition, with one exception; here subjects were exposed to only
names. Participants in the odor-name condition were presented with
both an odor and its congruent name. Here, the participants were
asked to retrieve autobiographical events for each pair of items. The
participants were instructed to direct an equal amount of attention
to both of the cues. The respective odor name was presented on the
computer screen in conjunction with the presentation of the odor.
The rating procedure in the odor-name condition was the same as
in the name-only and odor-only conditions, with the exception that
the subjects were exposed to both the odor and its congruent name.
Across conditions, thirty seconds were allowed for retrieval. When
all cues had been presented the memories were dated according to
the participants’ age at event.
Importantly, each participant received 20 stimuli and the evoked
memories were rated immediately after retrieval. For half of the cues
(n 10), the ratings were done in presence of the respective retrieval
cue, and for the other half (n 10) the ratings were done without
cues. The two rating formats were counterbalanced across partici-
pants in order to prevent potential effects of presentation order. In
addition, each participant was given a unique randomized item pre-
sentation order. The rating procedure was followed by distribution of
the questionnaire concerning health status and sensory aptitude.
The Number of Memories
The number of evoked memories across the three cue
conditions (i.e., odor only, name only, odor name) was
analyzed with a two-way ANOVA with cue condition and
gender as between-groups factors. The results from the
ANOVA showed that the number of elicited memories var-
ied over cue type [F(2,66) 3.10, p .05]. Specifically,
Tukey post hoc comparisons showed that the name-only
(M 9.46, SD 4.05) and odor-name (M 10.79, SD
4.20) conditions triggered equal number of memories, but
more than the odor-only condition (M 7.58, SD 4.26).
The effect of gender was not significant (F 1).
of olfactory information, suggesting that odor recognition
may be a joint product of perceptual, semantic and episodic
information (Larsson, 1997; Larsson & Bäckman, 1997).
Given that autobiographical odor memory draws on epi-
sodic memory, it is of interest to explore whether verbal
processes also influence the recollection of autobiograph-
ical olfactory information. In the present work, this topic
was examined by cuing subjects with odors in conjunc-
tion with their respective names, as contrasted with two
control conditions where subjects only received names or
odors. In this way, we could evaluate the potential effects
of cue format on age distributions and phenomenological
experience of autobiographical memories.
An additional aim was to investigate vantage point selec-
tion in autobiographical odor memory. Recent research has
highlighted two different modes of remembering referred
to as field and observer vantage points (Georgia & Neis-
ser, 1996; McIsaac & Eich, 2002). The former refers to a
recollection in which subjects envision the event as if see-
ing it again with their own eyes. In the latter, persons take
the perspective of a detached spectator (Georgia & Neisser,
1996). Field memories have shown to comprise more af-
fective reactions and physiological states, whereas observer
memories include more information about physical appear-
ance and object locations (McIsaac & Eich, 2002). Given
that vantage point selection is totally unexplored in autobio-
graphical odor memory we wanted to highlight this experi-
ential factor and its relation to olfactory cued memories.
Thus, the main aims of this study were to investigate
whether (1) verbal and conceptual processing influence
the age distribution and phenomenological qualities of ol-
factory evoked memories; and to elucidate whether (2) the
experiential qualities of retrieved information are influ-
enced by olfactory perception.
Seventy-two healthy volunteers (22 men and 50 women) between
65–80 years (M 71.69, SD 4.40) participated. An older age
cohort was chosen because of the possibility to investigate whole life-
span memory distributions. The participants were recruited through
senior citizen organizations in the Stockholm and Uppsala areas. All
participants reported being in good health as indicated by their self-
rated health (M 3.69, SD .88) (5-point scale where 1 very poor
and 5 very good). Also, participants rated their visual, auditory, and
olfactory aptitude on a 5-point scale where 1 very poor and 5
very good. The specific questions posed were: How good or bad is
your vision/olfaction/hearing as compared with other persons of your
own age? All subjects had sufficient visual (M 3.11, SD 1.0), au-
ditory (M 3.21, SD .90), and olfactory (M 3.14, SD 1.1) ca-
pacities in order to manage the sensory demands of the experiment.
Twenty stimuli were used as test materials and assigned to one of
two test sets. Each set consisted of 10 items. Based on previous data
(Willander & Larsson, 2006) the two sets of items were constructed
so that they would potentially elicit an equal amount of memories. The
two test sets comprised the following items: (1) tar, glühwein (mulled
wine), clove, whisky, snuff, black currant, chlorine, tobacco, bitter
almond, violet; and (2) salubrin (antiseptic), soft soap, cinnamon, lini-
ment, red wine, soap, lily of the valley, anise, cardamom, beer.
All odorants were kept in nontranslucent glass jars and covered
with cotton pads to prevent visual inspection. The odors were reg-
The Distribution of Memories Over
the Life Span
For each participant, the number of evoked memories
dated to a specific decade was divided by the subject’s total
number of memories. The proportions were then submitted
to a mixed two-way ANOVA with cue type as between-
groups factor, and decade as within-group factor. The
analysis yielded a main effect of decade [F(7,63) 12.18,
p .001]. Post hoc testing showed that a higher proportion
of memories were located to the first two decades as com-
pared to the third to eight decades. Also, more memories
were overall localized to the first than to the second decade.
Interestingly, the interaction between cue type and decade
was significant [F(14,126) 1.76, p .05]. The source
of this interaction was related to the first decade where
more memories were generated in the odor-only condition
as compared to the name-only condition. The amount of
memories generated in the odor-name and odor-only con-
ditions, and odor-name and name-only conditions did not
differ reliably (ps .05). No other comparisons were reli-
able. The age distributions across the three cue conditions
are displayed in Figure 1. Given the unbalanced number of
men and women across the three conditions meaningful
gender analyses could not be performed.
The experiential data for the evoked memories across
the two conditions (rating with the cue/rating without the
cue) were submitted to a 3 (cue type) r 2 (rating condi-
tion) mixed ANOVA. The first factor varied between sub-
jects and the second within subjects.
The ANOVA on rated pleasantness (valence) revealed a
main effect of cue type [F(2,69) 5.18, p .01]. Tukey
post hoc analysis indicated that odor-only evoked memories
(M 6.43, SD 1.74) were rated as more pleasant than the
name-only evoked memories (M 5.49, SD 1.38). The
memories evoked by odor names (M 5.97, SD 1.11) did
not differ from odor-only or name-only evoked memories on
rated pleasantness. The main effect of rating condition was
not significant (p .70). No interaction effect between cue
type and rating condition was observed (p .30).
A main effect of cue type was demonstrated for emo-
tionality [F(2,69) 3.59, p .05]. Tukey post hoc com-
parisons indicated that memories evoked in the odor-only
condition were more emotional (M 5.22, SD 2.00) as
compared to the name-only (M 4.01, SD 1.54) and
odor-name (M 4.23, SD 1.80) conditions. The name-
only and odor-name conditions did not differ reliably. No
main effect of rating condition was observed (p .20).
The interaction between cue type and rating condition was
not significant (ps .90).
Cue type had a reliable effect on the feeling of being
brought back in time to the occurrence of the event
[F(2,69) 4.30, p .05]. The odor-only evoked memo-
ries (M 6.66, SD 1.39) were experienced with a stron-
ger feeling of being brought back in time as compared
with memories evoked by name-only (M 5.68, SD
1.28) or odor-name cues (M 5.91, SD 1.61). Ratings
in the name-only and odor-name conditions did not dif-
Proportion of Memories
Age at Event (years)
0–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80
Proportion of Memories
Age at Event (years)
Proportion of Memories
Age at Event (years)
0–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80
0–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80
Figure 1. The distribution of (A) odor-only-, (B) name-only-,
and (C) odor-name-evoked autobiographical memories across
the life span. Error bars indicate 1 standard error.
tically different from the age distributions obtained for nei-
ther the name-only or odor-only cue conditions. A similar
pattern was observed for the phenomenological qualities,
such that perceived pleasantness, emotionality, and feel-
ings of being brought back in time were lower when odors
were presented with their respective names as compared
to when only an odor was presented. Indeed, when par-
ticipants were aware of the odor’s identity, the experiential
ratings did not differ from the ones obtained for names.
These observations suggest that olfactory knowledge pro-
duces a shift from a perceptual to a more conceptually
driven retrieval process (cf. Herz, 2000, 2003). Given that
the present findings suggest that different types of sensory
cues recruit different types of retrieval processes, it is of
interest to consider postulated retrieval models of auto-
biographical memory information. Notably, Conway and
Pleydell-Pearce (2000) distinguished between two different
retrieval processes of autobiographical memory: strategic
and automatic. In strategic retrieval, cues entail an inten-
tional cyclic and elaborative search process of information
in memory until a specific memory is formed, a process
that has been associated with verbal cues (Haque & Con-
way, 2001). In automatic retrieval, recollection is direct
and effortless and immediately activates a representation
of an event in memory. To the extent that odors presented
with labels primarily were experienced as verbal items,
the present observations suggest that participants in the
odor-name condition engaged in a more strategic retrieval
process, whereas, a presentation based on odors alone may
have induced a more direct and automatic activation of
sensory-specific autobiographical information (Conway &
Pleydell-Pearce, 2000). Moreover, the analysis on the num-
ber of memories triggered over the three different cue types
showed that more memories were generated when a verbal
cue was present (i.e., the name-only and odor-name condi-
tions) and fewer memories were accessed when only an
odor served as a retrieval cue. This outcome suggests that
fer reliably (p .90). Rating condition did not reach the
level of significance and the interaction effect between
cue type and rating condition did not influence the feel-
ings of being brought back in time (ps .20).
Regarding vividness no main effect of cue type or inter-
action effect between cue type and rating condition were
statistically reliable (ps .30). However, a main effect of
rating condition was found for vividness [F(1,69) 4.76,
p .05] such that rating with the cue (M 6.65, SD
1.51) produced higher vividness ratings as compared to
ratings without the cue (M 6.33, SD 1.61).
Finally, for the intensity ratings no reliable main effects
of cue type or rating condition (ps .30) were observed.
Also, the interaction was not reliable (p .10).
The vantage point proportions were submitted to a two-
way mixed ANOVA with cue type (i.e., odor only, name
only, and odor name) as between-group factor and vantage
point (i.e., field, observer, or neither) as within-group fac-
tor. The mean proportions of the vantage point ratings are
displayed in Table 1. The ANOVA showed a main effect of
vantage point [F(2,68) 114.77, p 001]. Post hoc test-
ing indicated that more memories were experienced with a
field perspective rather than an observer perspective, that
in turn were more common than events categorized as nei-
ther. The interaction effect was not significant (p .80).
In agreement with previous findings, the present work
suggests that odor-evoked memories are different from
other memory experiences. Events evoked by olfactory
information were older than memories associated with
verbal information (Willander & Larsson, 2006). Also,
odor-evoked memories were experienced as more emo-
tional and pleasant, and associated with stronger feel-
ings of being brought back in time as compared to events
evoked by verbal cues (Chu & Downes, 2000; Herz &
Cupchik, 1995; Herz & Schooler, 2002). Of more interest
however, were the observations that the age distribution,
the phenomenological qualities, and the number of evoked
memories were affected by explicit knowledge of the pro-
vided odor cue. This outcome suggests that conceptual
processes may have a significant influence on the retrieval
of autobiographical olfactory information.
Specifically, when odors were presented together with
its congruent name, the age distribution of memories for
the first three decades was significantly affected such that
the memory peak in childhood observed for odors only was
attenuated. In fact, knowledge of an odor’s name resulted
in a distribution taking an intermediate position; not statis-
Tabl e 1
Mean Proportions (With Standard Deviations) of Vantage
Points As a Function of Cue Type
Odor Only Name Only Odor Name
Vantage Point M SD M SD M SD
Field .73 .26 .78 .24 .69 .29
Observer .20 .23 .17 .22 .24 .25
Neither .07 .19 .05 .07 .07 .22
Mean Experimental Rating
Pleasantness Brought Back Emotionality
Figure 2. Experiential ratings across cue type. Error bars indi-
cate 1 standard error.
nomenal characteristics of autobiographical memories for positive, neg-
ative, and neutral events. Applied Cognitive Psychology, 17, 281-294.
Distel, H., & Hudson, R. (2001). Judgement of odor intensity is influ-
enced by subjects’ knowledge of the odor source. Chemical Senses,
Ehrlichman, H., & Bastone, L. (1992). Olfaction and emotion. In
M. J. Serby & K. L. Chobor (Eds.), Science of olfaction (pp. 410-438).
New York: Springer.
Georgia, N., & Neisser, U. (1983). Point of view in personal memories.
Cognitive Psychology, 15, 467-482.
Haque, S., & Conway, M. A. (2001). Sampling the process of auto-
biographical memory construction. European Journal of Cognitive
Psychology, 13, 529-547.
Herz, R. S. (1998). Are odors the best cues to memory? A cross-modal
comparison of associative memory stimuli. In C. Murphy (Ed.), Ol-
faction and taste XII: An international symposium (Annals of the New
York Academy of Sciences, Vol. 855, pp. 670-674). New York: New
York Academy of Sciences.
Herz, R. S. (2000). Verbal coding in olfactory versus non-olfactory cog-
nition. Memory & Cognition, 28, 957-964.
Herz, R. S. (2003). The effect of verbal context in olfactory perception.
Journal of Experimental Psychology: General, 132, 595-606.
Herz, R. S. (2004). A naturalistic analysis of autobiographical memories
triggered by olfactory visual and auditory stimuli. Chemical Senses,
Herz, R. S., & Cupchik, G. C. (1995). The emotional distinctiveness of
odor-evoked memories. Chemical Senses, 20, 517-528.
Herz, R. S., Eliassen, J., Beland S., & Souza, T. (2004). Neuroim-
aging evidence for the emotional potency of odor-evoked memory.
Neuropsychologia, 42, 371-378.
Herz, R. S., & Schooler, J. W. (2002). A naturalistic study of autobio-
graphical memories evoked by olfactory and visual cues: Testing the
Proustian hypothesis. American Journal of Psychology, 115, 21-32.
Herz, R. S., & von Clef, J. (2001). The influence of verbal labeling on
the perception of odors: Evidence for olfactory illusions? Perception,
Larsson, M. (1997) Semantic factors in episodic recognition of com-
mon odors in early and late adulthood: A review. Chemical Senses,
Larsson, M., & Bäckman, L. (1997). Age-related differences in epi-
sodic odour recognition: The role of access to specific odour names.
Memory, 5, 361-378.
Lehrner, J., Marwinski, G., Lehr, S., Johren, P., & Deecke, L.
(2005). Ambient odors of orange and lavender reduce anxiety and
improve mood in a dental office. Physiology & Behavior, 86, 92-95.
McIsaac, H. K., & Eich, E. (2002). Vantage point in episodic memory.
Psychonomic Bulletin & Review, 9, 146-150.
Packard, M. G., Cahill, L., & McGaugh, J. L. (1994). Amygdala
modulation of hippocampal-dependent and caudate nucleus- dependent
memory processes. Proceedings of the National Academy of Sciences,
Rubin, D. C. (1982). On the retention function for autobiographical
memory. Journal of Verbal Learning & Verbal Behavior, 21, 21-38.
Rubin, D. C., Groth, E., & Goldsmith, D. J. (1984). Olfactory cuing
of autobiographical memory. American Journal of Psychology, 97,
Rubin, D. C., Rahhal, T. A., & Poon, L. W. (1998). Things learned
in early adulthood are remembered best. Memory & Cognition, 26,
Rubin, D. C., & Schulkind, M. D. (1997). Distribution of important
and word-cued autobiographical memories in 20-, 35-, and 70-year-
old adults. Psychological Aging, 12, 524-535.
Savic, I., Gulyàs, B., Larsson, M., & Roland, P. (2000). Olfactory
functions are mediated by parallel and hierarchical processing. Neu-
ron, 26, 735-745.
Willander, J., & Larsson, M. (2006). Smell your way back to child-
hood: Autobiographical odor memory. Psychonomic Bulletin & Re-
view, 13, 240-244.
(Manuscript received April 19, 2006;
revision accepted for publication January 12, 2007.)
verbal cues overall are more efficient reminders of past
experiences. Taken together, the overall pattern of findings
suggests that retrieval of olfactory evoked information is
sensitive to semantic and conceptual processing.
The results of this study support the notion that olfac-
tory evoked memories are more emotional than memories
evoked by verbal cues (Chu & Downes, 2002; Herz, 2004;
Herz & Cupchik, 1995). As noted above, evidence is
equivocal regarding whether it is the memory by itself that
is more emotional or whether it is the perceptual proper-
ties of the olfactory cue that produce the higher emotional
arousal. In the present work, reexposure of the target cue
during rating did not significantly alter the phenomeno-
logical evaluations of the retrieved events. This observa-
tion suggests that olfactory evoked memories themselves
are associated with a higher emotional response and that
the perceptual stimulation alone cannot account for the
higher emotional arousal.
As noted above, recent work has highlighted two differ-
ent modes of remembering denoted field and observer van-
tage points. In congruence with other research addressing
vantage point selection in autobiographical memory, our
findings indicated that retrieved events, irrespective of cue
type, were experienced from a field rather than from an
observer perspective (D’Argembeau, Comblain, & Van der
Linden, 2003). Also, taking a field rather than an observer
perspective has been associated with more affective and
emotional reactions. Given that vantage point selection did
not vary across the different cue types, the present results
may reflect a general emotional tone in recollected events.
Taken together, the results of this study indicate that
retrieval of olfactory evoked information is sensitive to
semantic and conceptual processing. Specifically, the
age distribution, the phenomenological qualities, and the
number of evoked memories were significantly affected
by explicit knowledge of a provided odor cue. Also, the
outcome of this study supports the notion that olfactory
evoked memory representations are more emotional than
memories triggered by verbal information.
This work was supported by Swedish Research Council Grant
F0647/2001 to M.L. The authors thank Mats J. Olsson and Fredrik Jöns-
son for Uppsala lab space. Also, hearty thanks to all of the devoted par-
ticipants in the Autobiographical Memory Project II. Correspondence
concerning this article should be addressed to J. Willander, Stockholm
University, Department of Psychology, Frescati Hagväg 14, SE-106 91
Stockholm, Sweden (e-mail: firstname.lastname@example.org).
Chu, S., & Downes, J. J. (2000). Long live Proust: The odor-cued auto-
biographical memory bump. Cognition, 75, B41-B50.
Chu, S., & Downes, J. J. (2002). Proust nose best: Odors are better cues
of autobiographical memory. Memory & Cognition, 30, 511-518.
Conway, M. A., & Haque, S. (1999). Overshadowing the reminiscence
bump: Memories of a struggle for independence. Journal of Adult
Development, 6, 35-44.
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction
of autobiographical memories in the self-memory system. Psycho-
logical Review, 107, 261-288.
D’Argembeau, A., Comblain, C., & Van der Linden, M. (2003). Phe-