Olfactory Memory in Humans

Olfactory Memory in Humans
from The Temporal Lobe and Limbic System by Pierre Gloor, pages 316-318

[Entered by hand, likely many typos]

In humans, odors are reputed to have a particularly potent ability to evoke
old memories of events or situations that usually occured in a socioaffective
context of which that odor had been a part.  Memories evoked by such odors are
said to have an immediacy and vividness that is rarely equalled by those
evoked by other sensory stimuli.  I personally had an experience of this kind:
Upon entering a room at a friend's country house I was struck by an odor that
I immediately recognized as that of my grandmother's kitchen, into which I had
not set foot since my childhood.  The odor suddenly transported me back in
time and space to that kitchen with an overwhelming feeling of familiarity.
It was a place that under ordinary circumstances I thought about only on rare
occasions, and these recollections never evoked this feeling of being there.
Also, the odor would not have been accessible to my voluntary recall.
Anecdotal reports like this illustrate two striking features peculiar to the
olfactory sense: the durability of olfactory memories, which contrasts with
the common inability to reactivate them voluntarily, and the associative power
of the olfactory sense.

The most famous and often cited example of such a vivid recall of an old
memory by an olfacto-gustatory sensation is that described in the first volume
of Marcel Proust's novel A la Recherche du Temps Perdu (Remembrance of Things
Past).  Even though this is a literary and not a scientifically documented
example it is worth reviewing, because it bears such an uncanny resemblance to
the experiential response elicited by the electrical stimulation of the
amygdala that is described in Chapter 1 of this book.  In Proust's novel the
narrator relates an experience elicited by the olfacto-gustatory sensation of
tasting a certain kind of cookie, "une petite madeleine" dunked in tea that
was offered to him by his mother.  As soon as he put it into his mouth he was
seized by a strong but isolated delicious feeling of pleasure that permeated
him without any notion of what its source was ("un plaisir délicieux m'avait
envahi, isolé, sans la notion de sa cause")--a type of "tip of the tongue"
phenomenon that, in spite of its vividness, was initially impervious to any
attempt at identifying its origin.  In some way the flavor seemed to be
connected to a visual memory that was trying to surface in his consciousness,
and then suddenly that memory emerged with an overwhelmingly vivid clarity: it
was that of an event in the narrator's childhood, when while visiting his aunt
Léonie in the little town of Combray, she had offered him a "petite madeleine"
that she had soaked in her tea.  And with this memory there resurfaced a most
vivid visual recall of his aunt's house, which like a stage set became
superimposed upon the present scene (which had in the meanwhile been altered),
together with the memories of the little town of Combray, of some of the
places there, and of what he used to do there as a child.  The narrator makes
the point that he had many times seen "petites madeleines" in pastry shops and
that the mere sight of them, there or even now in the incident he describes,
had never evoked any memory.  It was only when he tasted the madeleine and
experienced its flavor--its odor and taste--that the memory surfaced in his
mind.  The sequence described by Proust bears a striking resemblance to that
elicited by electrical stimulation of the amygdala in the epileptic patient,
which is described in some detail in Chapter 1.  There, as in Proust's
literary example, the evocation of a remote childhood memory was at first
unfocused yet strong, with a "tip of the tongue" quality, and upon
restimulation there emergged the memory of reliving the old childhood event
with a startling vividness and immediacy.

Proust's description and those of similar experiences of odor-evoked memories
reported by others are thus in many ways similar to those some temporal lobe
epileptics give of the experiential hallucinations that usher in their
seizures: What characterizes this Proustian phenomenon is a sudden
overwhelming and compellingly vivid quality that calls forth memories coupled
with a feeling of reliving a past experience that is utterly inaccessible to
voluntary recall.  Are thus odors capable of tapping into the same memory
evocation mechanism that is activated in some tmporal lobe epileptics by
temporal lobe seizure discharge or by electrical stimulation of temporal lobe
structures?

A scientific approach to olfactory memory cannot, however, rely on such
anecdotal evidence.  An attempt has been made to document objectively the
power of odors to evoke vivid memories of past events and situations (Herz and
Cupchick, 1992).  College-student volunteers of both sexes were presented with
a series of odors and asked to describe any memories that were evoked by any
of them.  This occurred commonly and such memories frequently had emotional
connotations; were particularly clear; were usually not about events
frequently recalled by the subjects in their everyday life; and were
comparatively old, most often going back to early childhood.  This study thus
lends some experimental support to the notion that odors have the power to
evoke vivid memories from a remote past.  However, it can be criticized on the
grounds that only the effects of olfactory stimuli were studied, and thus no
comparison is possible with the capacity of visual or auditory stimuli to
evoke memories.

Other experimental studies have, however, shown that olfactory memories have
some peculiarities that indeed set them apart from memories in other sense
modalities.  Testing of human memory--for instance, in the visual
modality--often relies on verbal labeling of the memoranda.  Yet, in the study
of olfactory memories, naming is an inadequate tool.  People in a laboratory
testing situation have a poor ability to attach verbal labels to odors, even
to those with which they are familiar.  Although they readily assent to the
name of a familiar odor when it is supplied to them by the experimenter, they
are perplexed that they had been unable to think of it themselves (Engen,
1987).  Odors thus do not seem to stand on their own as individual,
well-circumscribed percepts, in contrast to visual objects and sounds.  This,
besides accounting for the naming difficulty, may explain why it is difficult
to conjure up in one's mind a well-known odor, whereas we have no difficulty
in imagining visual objects or sounds.  These peculiarities of olfactory
memory are probably explainable by the fact that in the olfactory system,
unlike in the visual system, elementary perceptual attributes are not treated
separately from global attributes.  There is nothing fundamentally different
in the way the perceptual attributes of a single chemical and those of a
mixture of chemicals are represented in the olfactory cortex: right from the
entry stage into the central nervous system olfactory stimuli are represented
non-topographically by a widely distributed matrix.  Thus, a perception caused
by a mixture of odors cannot be broken down into its simpler constituent
elements as is the case, for instance, in the visual system.  Herein lies on
of the strengths, but also one of the weaknesses, of the olfactory sense.
Even though by representing odors in "gestalt" form as distributed matricies
right from their port of entry into the brain, the olfactory system is capable
of representing and discriminating myriad different smells by very economic
means, it is weak in analytical power.  It cannot break down complex odors
into its constituents.  Odors seem to acquire significance for people in
relation to the context in which they experience them, but not in relation to
an abstract semantic label such as a name.  People are thus similar to animals
in this respect: odors acquire their significance through contextual cues.  As
discussed earlier, in macrosmatic animals, odors--in constrast to vemeronasal
stimuli--commonly have no innate behavioral significance; they acquire it
readily, however, through association, particularly in behavioral contexts
that are highly significant to the animal.

When olfactory memory is tested in humans with methods that utilize odor
recognition rather than naming, olfactory memory differs in two respects from
visual memory: its acquisition is relatively poor as compared with visual
memory, but its staying power by contrast is remarkable.  Thus, at very short
intervals--where the recognition of visual stimuli presented to a normal
subject is nearly 100%--memory is poorer for odors, where at best it reaches
90% and frequently is between 70% and 85% (Engen et al, 1973; Lawless and
Cain, 1975).  But visual recognition memory fades fast and drops to chance
levels within 3 months (Shepard, 1967); olfactory recognition memory shows ony
a slight decay and then remains on a plateau of 70% for months or even longer
(Engen and Ross, 1973; Lawless and Cain, 1975; Engen et al, 1991).
Familiarity, pleasantness or unpleasantness of the odor, or whether it can be
labeled by the subject, as well as the retention interval, has no influence on
the persistence of the memory.  This extraordinary staying power of olfactory
memory coupled with its strong contextual associative properties probably lies
at the root of the Proustian phenomenon.  The phylogenetic history of
olfaction suggests that this sense modality has been singled out in early
mammalian evolution to subserve the role of relating olfactory stimuli to
experiences that are highly significant to the animal in such behavioral
situations as, for instance, procreation and various forms of social
intercourse.  Olfactory stimuli quickly form strong associaitions with such
highly significant individually experienced events as shown by examples cited
earlier.  Such associations are very enduring.  This role of olfaction may be
less obvious in humans, but it survives.

There are probably nonolfactory memories that are as unforgettable as
olfactory ones.  They do not relate to isolated visual or auditory percepts,
but to stimuli contextually imbedded in the experience of highly meaningful
events of an individual's past.  These are often either highly emotional
memories or memories of childhood going back to a time when the fundamental
experiences that shape one's socioaffective connections to the social nexus in
which we are imbedded are formed, and are often concerned with kinship and
territoriality (the home or other places that have a "territorial flavor"
efining one's social nexus) and to frightening antagonistic encounters.  Such
memories are not necessarily limited to childhood experiences.  Memories of
this kind tend to be very enduring and seem to be as indestructible as
olfactory memories.  They are part of what defines our identity to ourselves.
These are the type of memories that are reactivated in some temporal lobe
epileptics by epileptic discharge or electrical stimulation of the temporal
lobe structures, mostly the limbic structures of the temporal lobe.  Olfactory
stimuli and complex constellations of nonolfactory stimuli occurring in
meaningful socioaffective contexts probably form enduring memories through
interactions with both the amygdaloid and hippocampal systems.  Reactivation
of such memories by re-experiencing the underlying perceptual context of a
fragment of it (particularly if that fragment is an olfactory stimulus) has
the power to reactivate such memories, sometimes as in Proust's example (and
replicated by the results of amygdaloid stimulation in the patient described
in Chapter 1) by arousing its affective component first.  Observations on the
evocations of such memories by electrical stimulation suggest that the
structure that is most prone to trigger such recalls in the amygdala (Feindel
and Penfield, 1954; Gloor et al, 1981, 1982; Fish et al, 1993).  From a
comparative biological perspective this appears understandable because the
amygdala is the locus where in macrosmatic animals the simplest associations
of this kind--those between behaviorally meaningful pheromonal stimuli
mediated by the vemeronasal system and olfactory stimuli encoding individual
life experience mediated by the main olfactory system--take place.