Chem. Senses 24: 317-325,
1999
© Oxford University Press
Odors: Implicit Memory and Performance Effects
Centre Européen Des Sciences Du Goût, Dijon, France
Correspondence to be sent to: Joachim Degel, Allmendring 12, D-75203 Königsbach, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In order to assess the influence of odors on human performance and implicit memory for odors, 108 subjects completed a variety of tests in weakly scented (jasmine, lavender or odorless) rooms without having been made aware of the odor. After a 30 min interval the subjects were shown slides of different surroundings, including the room they had been in, and were requested to rate how well a set of 12 odors, including a blank, would fit to these surroundings. Half of these contexts contained visual cues related to two of the presented odors (leather and coffee). After the rating of fit the subjects had to rate the odors for pleasantness, were asked to identify the odors with their correct names and to tell where and when they had last smelled these odors. One subject remembered smelling the odor (jasmine) in the room and was discarded from the analysis of the results for the rating of fit. None of the others reported recollection of the experimental odors. The results showed that in general jasmine had a negative and lavender a positive effect on test performance. If an odor-related visual cue was present in the context, the related odor was always rated highest in fit to that context. Furthermore, the subjects working in rooms with an odor subsequently assigned this odor to the visual context of that room to a significantly higher degree than subjects working in rooms with different odors. Since none of the subjects reported that they had smelled the odor in the rooms where performance testing took place, it was concluded that the memory for these odors was implicit. Further analysis showed that such memory was only found in subjects who were unable to supply the right name for the odor. The possible consequences of this latter finding for understanding the relationship between sensory (episodic) and semantic odor memory are discussed.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Influence of odor on human behavior and implicit odor memory are both phenomena that are generally accepted, notwithstanding the lack of scientific proof regarding their existence. Although there is a wide belief in aromatherapy (Tisserand, 1995
) and in
the fact that aromas can somehow magically influence behavior in everyday life, there is only
little systematic work on these issues compared with the amount of research for other modalities
(Schacter, 1987; Smith, 1989). In studies of
performance facilitation by ambient odors (Gilbert et
al.,1997)
or by merely suggested ambient odors (Knasko et al.,
1990) no
significant effects on performance measures like clerical coding or digit deletion tasks were
found. In both studies the subjects knew about the nature of the research and were explicitly
aware of the fact that the influence of ambient odors on subjects performance was examined. A
study by Torii et al. (1988) showed the
effects of several odors
on `contingent negative variation' (CNV) in the EEG. These effects were
interpreted as stimulative or sedative respectively. Despite the fact that direct facilitation of
performance as measured in reaction times was not found in this experiment, the assumption was
made that a moderate increase in the degree of arousal caused by a stimulating odor could have a
positive effect on performance rates, whereas a sedative reduction of arousal would have the
inverse effect. This assumption was based on the Yerkes–Dodson law, which describes
the relationship between arousal or vigilance and performance (Yerkes and Dodson,
1908; Wrisberg, 1994) as an inverted
U curve. According to
this law there is an optimal arousal level below and above which performance decreases. Below
this optimum a stimulating odor would raise performance and a sedative odor would decrease it.
Above the optimal arousal level (see later in discussion) these odors would have the opposite
effects.
In research on implicit memory for odors, a clear and systematic proof for or against this
form of human memory has not been produced so far. An experimental design used by Schab
and Crowder (1995) showed
`implicit' effects in the
olfactory modality at first, but was later unable to support these findings in subsequent
experiments. Furthermore, just like in the experiments about performance cited above, the
authors worked with explicitly or consciously presented odors in the learning phase of the
experiments. Degel and Köster (1998)
found evidence of implicitly
learned memories for odors in an experiment in which they asked the subjects to rate how well
odors fitted to visual contexts. In contrast to the work of Schab and Crowder (1995), the odors did not serve as explicit stimuli in the
learning phase. However, some
essential variables were not systematically controlled in this experiment, thus leaving room for
objections to the interpretation of their results.
In order to both explore the role of odors in facilitation of performance and to show a
systematic approach for detection of implicit memory in the olfactory modality this study
extends
the method suggested by Degel and Köster (1998). In the learning
phase, subjects were submitted to several performance tests under different odor conditions, but
without any information about the fact that odor was involved in the experiment. The tests were
chosen to measure creativity, concentration and mathematical abilities. Thirty minutes later, the
subjects were asked to rate the fit of several odors to different visual contexts, including the one
in which they had just worked, in order to show implicit memory effects in the ratings. After this,
the subjects had to rate the odors for pleasantness. Finally, in order to control for the absence of
effects of conscious perception on the results for the rating of fit, they were asked to label the
odors and to mention the place and time where they had last smelled the odor.
Based on the experiments by Torii et
al. and by Degel et al.,
there are two main hypotheses involved in the present study:
subjects in a stimulating odor condition are expected to be more creative and productive, and to
make less errors than those in the neutral or sedative odor condition. Also, if there is an influence
of implicit memory for odors, the rating of the fit between odors and environments will be
positively influenced by unconsciously or subliminally learned odors in the test rooms.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Participants
A total of 108 subjects, 54 men (average age 29.8 years) and 54 women (average age 30.1 years), participated in the study. The subjects were recruited by telephone and were paid FF 150 for their participation. The subjects were randomly split into three equivalent groups (C, J, L) of 36 subjects consisting each of 18 male and 18 female subjects (average age group C: 30.2 years, average age group J: 30.4 years, average age group L: 29.2 years).
Odors
In the `learning' phase jasmine and lavender were chosen as the odors in the
experimental rooms. According to Torii et al. (1988), jasmine is
supposed to be a stimulating and lavender a sedative odor. A non-odorized room served as the
control condition.
In the `rating of fit' phase, a set of 12 jars was presented, 11 of which
contained an odor at an equalized suprathreshold concentration, the other containing no odor at
all. Subjects were told that each jar contained an odor, although sometimes in such a weak
concentration that they might not smell anything. To avoid visual interference (Davis,
1981; Zellner and Kautz, 1990) the
jars were identical. They
were marked with a randomly chosen three-digit code. The position of the jars for the rating of fit
was balanced over all subjects.
The chosen odorants represented a variety of scents known from everyday life. Some of them were directly related to the visual cues shown in images of the rating of fit task, some had a perfume-like character and some (jasmine, lavender and no odor) had served as ambient odors in the testing rooms.Table 1 shows the odor qualities with a short description. Some of the results to be described later are also given in this table.
|
Visual materials
In the rating of fit phase, 12 pictures were used showing the three test rooms and different surroundings from everyday life. Three of them contained a visual cue for leather odor and three others contained a visual cue for coffee odor. The order of the pictures was kept constant over sessions. The sequence of the pictures was random ized with the restriction that pictures with and without a visual cue were alternated (seeTable 2).
|
Test material
In order to assess the subjects' performance a creativity test, a letter counting
concentration test and a mathematical test were used. The creativity test was based upon
open-ended associative thinking (Wallach and Kogan,
1965; Wyon et al.,1997), a
task in which subjects had to specify as many
instances of a given class of objects as they could think of. The classes of objects were (a) things
that are red, (b) things that will make noise, (c) things that are square and (d) things that move on
wheels. In the provided time for the creativity test (2 min) the subject was instructed to tell the
interviewer as many instances of these object classes as possible. To keep the subject
concentrated on the task and in order to avoid any confounding effects of differences in writing
speed on the results, the interviewer wrote down all the specified instances.
In the letter counting concentration test the subjects had to count the frequency of a specified letter in 25 rows consisting of 50 randomly arranged mirror image letters (b, d, p, q) each. After counting and indicating the correct frequency of letters in a row the subjects were instructed to go to the next row. A total of 8 min was provided for this test. The mathematical test consisted out of five blocks of 15 tasks. In each block there were five multiplication tasks (multiply a single digit number by a double digit number), five addition tasks (add a double digit number to a double digit number) and another five subtraction tasks (subtract a double digit number from a double digit number). After completion of one block the subject was instructed to go to the next block of tasks. A total of 8 min was allowed for this test.
Each subject was instructed and tested individually by an interviewer. The sequence of the tests was not varied over the subjects and interviewers.
Procedure
All the subjects of one group were assigned for performance testing to a room that was scented with either an ambient odor [jasmine (group J) or lavender (group L)] or no odor [control (group C)] respectively. The subjects were not told that odors played a role in the study or that odors were present in the rooms.
The subjects participated individually. After meeting the interviewer, the subject was taken to the test room corresponding to the group he/she was assigned to and made to wait for 15 min before the interviewer started the tests. The interviewers were instructed to keep a rigorous time schedule and to start the next test as soon as the time for the previous test had elapsed. The total duration of the tests, including the initial 15 min waiting time and instructions, was 45 min.
In order to control for interviewer effects, the three interviewers were rotated systematically over the three experimental groups. Thus, each interviewer saw an equal number of men and women in each group and in each test room. Moreover the interviewers were kept unaware of the fact that the three rooms had different odors. They were also unaware of the aim of the study. After completion of the three tests, the subjects were taken to a different room where they took part in an experiment on the hedonic properties of three kinds of cookies, which lasted for 30 min. At the end of this experiment the subjects took part in the final stage of the experiments: the rating of the fit between odors and environments which took place in still another room. The subjects were seated in groups of three, separated by side walls, in front of a screen on which the visual images of the different contexts were projected. They each had the set of 12 odor jars in front of them. The subjects were instructed to rate how well each odor fitted each of the contexts shown. Ratings of fit were made on a 100 mm visual analog scale with the end labels `does not fit' and `fits'. After rating the fit of all odors to a visual context, a new context was shown on the screen. To neutralize the odor perception and cause a temporal delay between the ratings of fit the subjects were told to smell the inside of their own arm after each rating. In order to reduce olfactory adaptation a pause of 45 s was made before a new visual context was shown.
After completion of the rating of fit and a pause of 5 min the subjects had to rate the 12 odors for pleasantness, and were asked to write down the name of the odor and the place and time where and when they had last smelled the odor.
Statistical analysis
The analysis was performed with SPSS for Windows, Version 6.1.3 (SPSS, Inc.,
1994).
For the letter counting and mathematical tests, measures of `Amount' (rate of
processed tasks per 100 test tasks) and `Error' (rate of errors per 100 processed
tasks) were calculated. The creativity score (C-score) was calculated according to Wyon et
al. (1997). The score for having mentioned a
given category was
calculated in accordance to information theory as log2(1/P) where
P
is the
probability that the answer is given by a randomly selected subject. This score can
range from zero if all subjects mention the same category (P = 1) to 6.75 (P =
1/108) if only one of all 108 subjects mentioned a given category. Category definitions were
determined by an independent judge in a blind procedure. The C-score for a given subject is the
sum of the scores to all the responses given by that subject. The C-total, each subject's
total (unweighted) number of accepted (right) and non-accepted (wrong) responses, was recorded
separately as a further criterion. Some subjects were discarded from the calculations because they
obviously did not seriously participate in the tests and as a result their scores were either
completely outside of the normal range or incomplete. In the first case the normal rule in SPSS
for discarding outliers was applied. These cases are indicated in the results section.
For the rating of fit, normalized means were calculated. Normalization was performed by taking the rating of fit for each individual odor and dividing it by the mean of the 12 contextual ratings for that same odor by the same subject. Thus, values below 1.00 express a rating below the average, values above 1.00 a rating above the average rating for that odor.
If a subject explicitly remembered the odor from the testing room, his data were excluded from the analysis. For the rating of pleasantness the means per odor were calculated over subjects. For odor identification only an exact definition of an odor's name was counted as a correct answer. Since for some odors of a more general nature (ambient odor) no exact names could be given, only the results of the well-defined odors were taken into account in further treatment.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Performance testing
Letter counting
The data were analyzed with two 3 x 3 ANOVAs (group C, J, L x interviewer
A, B, C). Three cases were excluded from the analysis due to the extremely high rates of Error.
There were no significant effects of group or interviewer on the measures of Amount or Error.
For the measure of Error this result is basically caused by the rather similar results in groups C
and J (C: mean = 0.27, SD = 0.14; J: mean = 0.27, SD = 0.15). Therefore, both these groups
were
combined for another 2 x 3 ANOVA (group C and J, L x interviewer A, B, C).
There was a significant effect of odor condition [F(1,99) = 4.59, P < 0.05]
on
the rates of Error. Further analysis showed that Error rates in the lavender condition (group L)
were lower than those in the combined conditions C and J (C and J: mean = 0.27, SD = 0.14; L:
mean = 0.21, SD = 0.12).
Mathematical test
The influence of odor condition and interviewer on Error and Amount rates were analyzed
with two 3 x 3 ANOVAs (group C, J, L x interviewer A, B, C). Ten cases were
excluded from the analysis because of the extremely high rates of Error. There was a significant
effect of odor condition [F(2,89) = 5.99, P < 0.01] on the Error rates but no
effect on the rate of Amount. Here, the subjects made significantly more errors in the jasmine
condition (group J) than in the two other conditions. (C: mean = 0.06, SD = 0.04; J: mean = 0.09,
SD = 0.06; L: mean = 0.05, SD = 0.03).
Creativity test
Eight cases were excluded from the analysis of the C-scores due to no response in one of the
four tasks; the application of the logarithmic scale was therefore not possible. A 3 x 3
ANOVA (group C, J, L x interviewer A, B, C) showed no significant effects of the factor
group on the results of the C-scores [F(2,91) = 2.069, P = 0.132]. However a
significant influence of the factor interviewer was found in the data [F(2,91) = 8.618,
P < 0.01]. Further analysis showed that interviewer A produced the highest C-scores
followed by interviewer C and B (A: mean = 20.57, SD = 6.55; B: mean = 14.49, SD = 4.88; C:
mean = 17.688, SD = 5.93).
A 3 x 3 ANOVA (group C, J, L x interviewer A, B, C) on C-total showed a main effect of the two factors [F(4,99) = 10.835, P < 0.01], where both factors had a significant influence [interviewer: F(2,99) = 14.523, P < 0.01; group: F(2,99) = 7.146, P < 0.01]. There was no interaction between the factors [F(4,99) = 0.954, P = 0.436].
Like for the analysis of the C-scores, interviewer A caused the highest scores followed by interviewer C and B (A: mean = 59.19, SD = 13.72; B: mean = 43.06, SD = 12.53; C: mean = 49.81, SD = 14.13). The C-total was highest in the condition C followed by condition L and J (C: mean = 57.0, SD = 16.41; J: mean = 45.97, SD = 13.6; L: mean = 49.08, SD = 12.62).
Further analysis indicates an influence of the condition on the performance of the interviewers B and C, whereas interviewer A produced a relative stable result over the three different conditions.Table 3 shows the means and SDs for each interviewer over all conditions and the results of a one-way ANOVA which was computed for each interviewer over the conditions (interviewer i x condition C. J, L).
|
Rating of fit
The results of the rating of fit for all subjects together are shown inTable 4. Only one subject was excluded from the analysis because the person explicitly remembered smelling the odor (jasmine) in `test room J'. This person was also omitted in the results ofTable 5 below. All other subjects had been unaware of the presence of odors in the test rooms. Inspection ofTable 4 shows that for these cases where a visual cue for the odor was present, the highest values are indeed found for this odor (coffee or leather), with the exception of the office with the cup of coffee. There, the rating of fit of the no odor control was equally high. It is also clear that, in cases where a visual odor cue was present, smaller relative differences between the mean and the median are found. This means that fewer subjects gave exceptionally high ratings than in the context without a visual cue. In general, it is remarkable that the means are in all cases higher than the medians. This cannot just be due to one or a few subjects, since the data are normalized per individual and therefore each individual average is 1.00 by definition. Furthermore, it is clear that in two contexts, the kitchen (with visual cue) and the train lavatory (without visual cue), odors were expected. They had the lowest values in the no odor column and both had high ratings of fit for most of the other odors, except for leather and laundry in the case of the kitchen and coffee and leather in the case of the lavatory, where the aftershave odor seemed to fit very well indeed. The three test rooms did not raise any special expectations with regard to odor. They had the highest fit values of all contexts in the no odor column (C = 1.24; J = 1.19; L = 1.34). The averages over all odors for them are not significantly different, although the one for room C (0.78) is somewhat lower than that of the other two (0.84 and 0.89), which are in the same range as those for the other contexts with the exception of kitchen and train lavatory.
|
|
Ratings of fit in the experimental groups and the control group
The results of the ratings of fit for the two test odors and the control odor to the three test rooms given by the three separate groups are shown inTable 5.
The influence of odor condition on the ratings of fit in the testing rooms was analyzed with a 3 x 3 ANOVA (group C, J, L x interviewer A, B, C) on the normalized magnitudes. For the visual context `testroom C' (no odor) and the rating of fit of the control odor there was a significant effect of group [F(2,99) = 3.92, P < 0.05]; C rated the control higher than group J and L (C: mean = 1.5, SD = 0.84; J: mean = 1.15, SD = 0.57; L: mean = 1.1, SD = 0.68). A subsequent t-test showed no difference in the scores of group J and L. In the ratings of fit for `testroom J' and the odor jasmine a significant influence of the factor group was also found, [F(2,98) = 3.67, P < 0.05]; here group J rated the odor of jasmine higher than group C and L did (C: mean = 0.55, SD = 0.46; J: mean = 0.98, SD = 0.45; L: mean = 0.64, SD = 0.82). A subsequent t-test showed no significant differences between group C and L. There was no significant influence on the ratings of fit for the context `testroom L' and the odor lavender, although the ratings for lavender were highest for this context (C: mean = 0.81, SD = 0.59; J: mean = 0.65, SD = 0.55; L: mean = 0.88, SD = 0.86).
Ratings of pleasantness
At the end of the experiment, the pleasantness of the odors was assessed (see Table 1) and the subjects were asked to identify the odors. Lavender was judged to be more pleasant than jasmine. Exposure to the odors in the rooms had no influence on the pleasantness ratings. The three groups did not differ in their ratings of pleasantness. There were no gender differences in the pleasantness judgements. Pleasantness of the odor had also no effect on the rating of fit. Of the two odors for which a visual cue was provided, coffee was judged to be more pleasant than leather
Visual cues, identification of the odors and the rating of fit
As regards identification, the odor of jasmine was correctly identified and labeled by only one subject, as were most of the other odors with the exception of the odors of lavender (56.5% correct), coffee (77.8% correct) and leather (47.2% correct). Using a more lenient criterion by allowing, for instance, wood as a more general descriptor than sandalwood or cedarwood did not change the results in a significant way. No gender difference in correct identification was found for any of these. Exposure to lavender had no influence on the correctness of the identification of this odor. All groups identified lavender equally well. The correct or incorrect identification of lavender had no influence on the results of the performance tests in any of the three groups. The relationship between the correct identification of an odor name and the ratings of fit is shown in Table 6 for the odors coffee and leather, in the situations where a visual cue was present and in one special case where a visual cue was absent. In all other situations without a visual cue there was no difference in the ratings of fit between correct and incorrect identifiers.
|
Whenever an odor-related visual cue was present in the picture, the rating of fit of the corresponding odor to the picture was the highest one found for that picture (seeTable 4). As could be expected, there was a clear relationship between the ability to identify the odor and the rating of fit in most cases where a visual cue was present (seeTable 6). In two of the six cases (train compartment and canteen room) the identifiers rated the fit not significantly higher than the non-identifiers. In the contexts without a visual cue no differences between identifiers and non-identifiers were found for the odors of either coffee or leather, with the exception of the context train lavatory. Here the non-identifiers for leather rated the fit of this odor significantly higher than the identifiers.
Odor exposure, identification of the odors and the rating of fit
In the condition lavender the correct identification of the odor name influenced the ratings of fit.Table 7 shows the means and Sds of the ratings of fit between test room L and lavender for correct and incorrect odor identifiers in the groups that were not exposed to lavender (C + J) and the group that was (L).
|
A 2 x 2 ANOVA (group C + J, L x identification correct, incorrect) shows no main effects, but a significant interaction [group x identification F(1,104) = 4.66, P < 0.05] is found. Further testing with one-way ANOVA showed a significant difference between the two groups for the ratings of fit of the incorrect identifiers. The incorrect identifiers in the lavender group rated the fit higher than the incorrect identifiers in the other groups. No such difference was found for the correct identifiers. These findings explain the fact that for the total group no significant proof of implicit memory was found for lavender inTable 4. The subjects who can identify this odor do not show such an implicit memory at all and therefore their results are responsible for the lack of significance found for the total group.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The first conclusion of this study is that unconscious perception of an odor can significantly affect the rate of errors made in the mathematical and letter counting tests. In the presence of the odor of lavender subjects made less errors than in the presence of no odor or the odor of jasmine. At first sight these results are contrary to the supposition of Torii et al. (1988
), who, on the basis of physiological measurements, supposed that jasmine had
a stimulating and lavender a sedative effect. However, if one supposes that performing in
psychological tests is a rather stressful experience, the results may be explained by the
Yerkes–Dodson law, which states that when the arousal level has passed a maximum, the
performance will decrease. If lavender is indeed a sedative stimulus and helps to reduce the
stressful arousal of the subject, it would lead to a better performance, whereas further arousal by
jasmine would have the opposite effect. Whether the same result would have been obtained
under normal unstressed conditions is doubtful. This means that one cannot base predictions
about the effects of odors on physiological measures such as variations in the CNV Torii et
al., (1988) only. The odors also had an
influence on the number of
valid responses in the creativity test, but not on the quality of these alternatives. Again the
subjects did less well in the jasmine condition than in the control condition, but this time the
difference between the jasmine and the lavender conditions was minimal. Furthermore, it could
be seen that the odor conditions influenced the numbers of responses written down by two of the
interviewers. Whether this was due to the utterances of the subject or to the behavior of the
interviewers cannot be decided on the basis of these results.
Secondly, the present experiment confirms the finding by Degel and
Köster (1998) that whenever people detect
visually a possible odor
source that
might be related to the odors they smell, they will make the connection (Table 6). It is quite understandable that they do so, especially when
they identify the odor
that belongs to the source. In a way this may also explain the exceptional case of the odor of
leather in the train lavatory, where the incorrect identifiers rated the fit between odor and
environment much higher than the correct identifiers. If the odor of leather is identified it does
not bear a special relationship with a toilet, but if it is not recognized as such it may be related,
because it has many animal and fecal undertones.
The third and perhaps most important conclusion of the study is that implicit memory for
odors truly exists. Not only is it clear that odors are expected more in some situations (kitchen
and train lavatory) than in others (Table 4), but
when a subject is exposed
unknowingly to an odor, he/she later connects this odor with the place he/she has been exposed
to it. This is true even when that person remains completely unaware of the fact that there ever
was an odor in the room in which he/she was tested. The most surprising result of this study is
that this implicit memory seems to work only in subjects that do not know the odor well enough
to identify it by its proper name afterwards. This was obviously the case in the control condition
where no odor was present, in the jasmine condition where none of the subjects (except one who
remembered having smelled it in the test room and was discarded from the analysis) could
identify the odor, and for the non-identifiers in the lavender condition. On the other hand, the
people who knew the name of lavender and identified it afterwards behaved like the subjects in
the two other groups. It should be stressed once more that none of these subjects remembered the
presence of lavender in the test room. When asked where and when they had smelled this odor
for the last time before the last session, none of them mentioned the room or even the laboratory.
Four possible explanations of this phenomenon are suggested. Two of these are based on the
supposition that the subjects did consciously perceive and recognize the odor, but subsequently
forgot that they had smelled it in the room. In the first place, this could simply be due to the fact
that they thought the recognized odor to be completely irrelevant with regard to the test situation
in which they were brought. In the second place it could be due to the retroactive inhibition
caused by the odors in the rating of fit procedure that preceded the identification question. In that
case the identifiers had identified lavender in the test room and had stored it in their memory, but
had forgotten it under the influence of the many odors (including lavender) that they were
exposed to in the rating of fit task. If true, this explanation would invalidate the supposition of
Engen (1991) and Herz and Eich (1995) that the
longevity of olfactory memory is due to the absence of retroactive inhibition. The third and
fourth possibilities would be that the presence of the name lavender in verbal memory blocks the
build-up of or the retrieval from the implicit memory. In the last case, it is supposed that the
subjects did not identify the odor in the test room and kept only an implicit memory of it, just
like the non-identifiers. When in the rating of fit session they were confronted explicitly with the
well-known odor of lavender, they simply denied the existence of this subtle memory, because
they felt that if the odor had been there they would certainly have noted it. This might also
explain why their average rate of fit was even somewhat (although not significantly) lower than
that of the correct identifiers of lavender in the other groups. The other possibility might be that
knowing an odor by name makes it impossible to build up new episodic memories for that odor,
even when one is not aware of its presence. This explanation is in line with the supposition of
Engen (1991) and Herz and Eich (1995) to a
certain extent. It might mean that once an odor has entered in semantic memory, this memory
takes over and prevents further interference by retroactive inhibition in episodic memory. The
identifiers do not store the new episodic odor experience. This would also explain why proactive
interference is so strong in olfactory memory (Lawless and
Engen, 1977; Engen, 1991).
Nevertheless, new episodic odor experiences are being
built up by the non-identifiers when no interference by semantic knowledge is involved. Whether
such new links do not cause retroactive inhibition in episodic memory remains to be seen. The
non-identifiers differ from the identifiers mainly through the fact that they cannot find the
connection between odor and name in semantic memory, because it is hardly imaginable that
they have never encountered lavender (or jasmine) in their life and heard the names of these
odors.
The fact that the train lavatory was the only case in which the non-identifiers of leather rated the fit of this odor significantly higher than the identifiers fits well into this last hypothesis, although the explanation offered above may apply equally well. Furthermore, if the toilet served as a visual cue, it could do so only for those who did not recognize the odor.
Whatever the truth of these hypotheses, which are now the subject of further research, the
accidental finding that the non-identifiers have a better implicit memory seems to be in
opposition with the general finding that knowing the correct name of an odor has a positive effect
in odor recognition experiments (Lyman and McDaniel,
1986, 1990). The question then arises
of what the differences between the two
methodological approaches to odor memory are that lead to such contrasting results. First of all it
should be noted that in odor recognition experiments the subject is asked explicitly to indicate
whether he/she has smelled the odor on a previous occasion, whereas in the research presented
here the presence of the odor in the learning phase was never made explicit. Furthermore, in the
learning phase of recognition experiments the odors are explicitly presented to the subjects, even
if they are unaware of the fact that they will later be asked to recognize them. If in these
situations the subjects label the odors with a name and thus enter them into semantic memory,
they have two ways of remembering them and it is not surprising that they do so better than
when
they do not find a name for them. That finding the right name leads to an even greater advantage
is also logical. This perhaps tells more about semantic memory than about odor memory.
|
Acknowledgments |
|---|
This study was granted in part by the German Savings Bank Association (Deutscher Sparkassen- und Giroverband), Bonn.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Davis, R.G. (1981) The role of non olfactory context cues in odor identification. Percept. Psychophys., 30,83 –89.[Web of Science][Medline]
Degel, J. and Köster, E.P. (1998) Implicit memory for odors: a possible method for observation. Percept. Motor Skills,86 ,943 –952. [Web of Science][Medline]
Engen, T. (1991) Odor Sensation and Memory. Praeger, New York.
Gilbert, A. N., Knasko, S.C. and Sabini, J. (1997) Sex differences in task performance associated with attention to ambient odor. Arch. Environ. Health, , 52, 195–199.
Herz, R.S. and Eich, E. (1995) Commentary and envoi. Implicit measures of odor memory. In Crowder, R.G. and Schab, F.R. (eds), Memory for Odors. Erlbaum, New York, pp. 159–175.
Knasko, S.C., Gilbert, A.N. and Sabini, J. (1990) Emotional state, physical well being, and performance in the presence of feigned ambient odor.J. Appl. Soc. Psychol. , 20,1345 –1357.
Lawless, H. and Engen, T. (l977) Associations to odors: interference, mnemonics, and verbal labeling. J. Exp. Psychol.,3 , 52–59 .
Lyman, B.J. and McDaniel, M.A. (1986) Effects of encoding strategy on long-term memory for odors. Quart. J. Exp. Psychol., 38 ,753 –765.
Lyman, B.J. and McDaniel, M.A. (1990) Memory for odors and odor names: modalities of elaboration and imagery. J. Exp. Psychol.: Learn., Mem. Cogn., 16,656 –664.
Schab, F.R. and Crowder, R.G. (1995) Implicit measures of odor memory. In Crowder, R.G. and Schab, F.R. (eds), Memory for Odors. Erlbaum, New York, pp. 71–91.
Schacter, D.L. (1987) Implicit memory: history and current status. J. Exp. Psychol., 13,501 –518.
Smith, A. (1989) A review of the effects of noise on human performance. Scand. J. Psychol., 30,185 –206.[Web of Science][Medline]
SPSS (1994) SPSS 6.0 für Windows: Anwenderhandbuch für das Base System [SPSS 6.0 for Windows: Handbook for the Base System]. SPSS Inc., Chicago, IL.
Tisserand, R. (1995) The Art of Aromatherapy. C.W. Daniel, London.
Torii, S., Fukuda, H., Kanemoto, H., Miyanchi R., Hamauzu, Y. and Kawasaki M. (1988) Contingent negative variation and the psychological effects of odour. In Van Toller, S. and Dodd, G.H. (eds), Perfumery. Chapman & Hall, London, pp. 107–120.
Wallach, M.A. and Kogan, N. (1965) Modes of Thinking in Young Children: a Study of the Creativity–Intelligence Distinction. Rinehart & Winston, New York.
Wrisberg, C.A. (1994) The arousal–performance relationship. Quest, 46,60 –77.
Wyon, D.P., Abrahamsson, L., Järtelius, M. and Fletcher. R.J. (1997) An experimental study of the effects of energy intake at breakfast on the test performance of 10-year-old children in school. Int. J. Food Sci. Nutr., ,48 , 5–12.[Medline]
Yerkes, R.M. and Dodson, J.D. (1908) The relation of strength of stimulus to rapidity of habit formation. J. Comp. Neurol. Psychol. ,18 , 459–482.
Zellner, D.A. and Kautz, M.A. (1990) Color affects perceived odor intensity. J. Exp. Psychol.: Hum. Percept. Perform., 16 ,391 –397.[Web of Science][Medline]
Accepted February 9, 1999
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. J. Bulsing, M. A.M. Smeets, and M. A. Van den Hout The Implicit Association between Odors and Illness Chem Senses, February 1, 2009; 34(2): 111 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Chrea, D. Grandjean, S. Delplanque, I. Cayeux, B. Le Calve, L. Aymard, M. I. Velazco, D. Sander, and K. R. Scherer Mapping the Semantic Space for the Subjective Experience of Emotional Responses to Odors Chem Senses, January 1, 2009; 34(1): 49 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Gaygen and A. Hedge Effect of Acute Exposure to a Complex Fragrance on Lexical Decision Performance Chem Senses, January 1, 2009; 34(1): 85 - 91. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Dematte, D. Sanabria, R. Sugarman, and C. Spence Cross-Modal Interactions Between Olfaction and Touch Chem Senses, May 1, 2006; 31(4): 291 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Sakamoto, K. Minoura, A. Usui, Y. Ishizuka, and S. Kanba Effectiveness of Aroma on Work Efficiency: Lavender Aroma during Recesses Prevents Deterioration of Work Performance Chem Senses, October 1, 2005; 30(8): 683 - 691. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Gedney, T. L. Glover, and R. B. Fillingim Sensory and Affective Pain Discrimination After Inhalation of Essential Oils Psychosom Med, July 1, 2004; 66(4): 599 - 606. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Holmes and C. Ballard Aromatherapy in dementia Adv. Psychiatr. Treat., July 1, 2004; 10(4): 296 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.A. Koster, J. Prescott, and E.P. Koster Incidental Learning and Memory for Three Basic Tastes in Food Chem Senses, June 1, 2004; 29(5): 441 - 453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Danuser, D. Moser, T. Vitale-Sethre, R. Hirsig, and H. Krueger Performance in a Complex Task and Breathing under Odor Exposure Human Factors: The Journal of the Human Factors and Ergonomics Society, January 1, 2003; 45(4): 549 - 562. [Abstract] [PDF]
|
|
|
|
|
|
H. N.J. Schifferstein and S. T. Blok The Signal Function of Thematically (In)congruent Ambient Scents in a Retail Environment Chem Senses, July 1, 2002; 27(6): 539 - 549. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. P. Koster, J. Degel, and D. Piper Proactive and Retroactive Interference in Implicit Odor Memory Chem Senses, March 1, 2002; 27(3): 191 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Degel, D. Piper, and E. P. Koster Implicit Learning and Implicit Memory for Odors: the Influence of Odor Identification and Retention Time Chem Senses, April 1, 2001; 26(3): 267 - 280. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||