902

SA MEDIESE TYDSKRIF 5 DESEMBER 1981

Review Article

Symptomatology of temporal lobe

epilepsy
V.M.NEPPE
Summary
The numerous and varied symptoms of temporal
lobe epilepsy are reviewed in' detail. Partial
elementary, partial complex, secondary
generalized, post-ictal and predisposing
aetiological symptoms are examined.
S. Atr. med. J., 60,902 (1981).

Slater and Roth l stress the immense variations in symptoms in

temporal lobe epilepsy. They also remark on how 'fixed and
stable' symptoms are in specific patients. Because of its protean
presentation, symptoms of temporal lobe epilepsy (TLE) are
difficult to classify. Broadly, they may have specific localizing
value to the temporal lobes, or they may reflect lesions in closely
allied areas such as the deep Sylvian insular area. As they are
epileptic in nature, these features may be paroxysmal and
sudden. On the other hand, there may be no specific localizing
features, for example in tonic-clonic seizures withour any aura
('warningless'). These would reflect a generalized seizure which
may be ab initio, or may be secondary to a very rapid partial
seizure, of which the patient may be amnesic.
Symptoms specifically referable to the temporal lobe and
sometimes found in TLE will be discussed here. All these
features may occur singly or in combination ('compound partial
seizures'). The occurrence of compound seizures greatly
increases the probability of localization to specific areas. For
example, intense fear is a nonspecific psychopathological feature
and its occurrence together with an unpleasant burning smell
suggests peri-amygdaloid localization.
The value of compound symptomatology in lesion localization
is stressed by Bingley.2 However, even the occurrence of a minor
specific local discharge, as in elementary partial seizures,
constitutes a 'small focal seizure', and as such must be construed
as an epileptic focal phenomenon. 3
TLE symptomatology is classified here in outline according to
Gastaut. 4 Wherever possible, sufficient validating evidence is
cited to substantiate the occurrence of the specific symptom in
the temporal lobe.

Elementary partial seizures of temporal

lobe origin
Ervin 5 points our that seizures occurring in isolation involve the
Department of Psychiatry, University of the Witwatersrand,
Johannesburg
V. M. NEPPE, M. B. B.CH., D.P.M., B.A., F.F. PSYCH (SA), M.MED. (PSYCH.),
PH.D. (MED.)
Date received: 19 December 1980.

'same abnormal firing of neuronal assemblies and have the same

significance for localization of pathology and for
neuropsychological theory'.
Despite the vas t integrative role of the temporal cortex and t~e
usual complexity of isolated seizures as a result of tliis, certam
seizures of simple, elementary quality do occur, possibly at the
primary cortical reception area. These symptoms may, and often
do, occur in isolation. Nevertheless, their localizing significance
cannot be ignored.
The demarcation into elementary and complex symptoms is
convenient bur artificial, 'and should be used flexibly. Many
sensations of the same modality can manifest at both elementary
or complex levels. To facilitate description they will be discussed
together under the technical heading of 'elementary partial
seizures' (as in the case of audition).

Elementary symptomatology with motor

symptoms
Gastaur 4 classifies aphasic and phonatory seizures under this
heading; although temporal lobe aphasia usually signifies a
receptive abnormality, it can be mixed.
Ictal aphasia is a common symptom of TLE; Hecaen and
Piercy (in Bingley2) found it in 37% of 97 right-handed and 90%
of 29 left-handed TLE patients. Many ofthese patients may have
speech and not genuine language disturbances, and the incidence
of receptive aphasia for the sample was about 13%. ThIS figure
compares with that obtained by Currie et al. 6 in the London
Hospital (1949 - 1967), based on clinical and EEG findings; they
found a 22% rate of speech disturbance in TLE patients. Bingley2
found an incidence of 48% with dominant hemisphere TLE and
21 % with bilateral foci; in his sample none of the patients with
non-dominant TLE had ictal aphasia. The site of origin of
temporal lobe aphasia can be in Wernicke's area.? Rasmussen s
mentions that those ictal aphasia elementary seizures originating
in the dominant parietotemporal speech zone (around
Wernicke's area) are often followed by numbness of the face or
arm.

Elementary symptomatology with special

sensory or somatosensory symptoms
Somatosensory phenomena classically derive from the
somatosensory cortex. s More complex somatosensory
phenomena derive predominantly from the parietal lobes, giving
rise to agnosias of various kinds. 5 Penfield 3 points out, however,
that somatic sensations referring to the limbs or body of either
side (more often contralaterally) may be obtained from the upper
bank of the Sylvian fissure and the insula as well, fitting into the
symptomatology for his 'temporal and deep Sylvian seizures'.
Penfield further stressed a group of poorly defined sensations
which at times derive from the anterior and mesial surfaces of the
temporal lobe. These may take the form of head sensations
(cephalic aura, the exact localizations of which are obscure) or

SA MEDICAL JOURNAL 5 DECEMBER 1981

body sensations, and may involve sensations of thermal quality

(heat or cold).3
Visual ictal features. RasmussenB divides visual illusions
and hallucinations into the categories of simple and complex, and
believes that elementary symptoms are limited to the occipital
lobe, involving spots or dots (coloured or black) or just 'lights'.
These may be static, moving or rotating. He regards complex
visual phenomena as deriving from the temporal lobe. \XTilliams 7
elaborates on the description of complex visual hallucinations of
temporal lobe origin; induction is frequently via a highly specific
precipitator, the hallucination may recur in a stereotyped way,
and it is almost always linked to other qualities of perception
such as voices, emotions or time. \XTilliams stresses, however,
that hallucinations of the special senses are uncommon in
temporal lobe lesions, and are of no further localizing value.
Bingley2 found visual hallucinations in 18% of dreamy states (the
same figure he found for auditory hallucinations). Currie ec al. 6
found that 18% (121) of his TLE sample had visual symptoms of
some nature. His corresponding figure for auditory symptoms
was 16% (107 cases). Blurred vision is also common, and was
found in 16% of Feindel and Penfield's 1954 sample of 50
patients with dreamy states. 9
Audi tory perceptual abnormalities. The transverse gyrus
of Heschl, situated in the middle of the superior temporal gyrus,
and the adjacent areas, passing inward to posterior to the island
of Reil, are devoted to crude auditory sensation. 3 Stimulation of
this area causes buzzing, hissing, ringing, thumping or
throbbing sounds. Rasmussen B mentions other kinds of simple
sounds, such as whistling, knocking, roaring and tapping, as
being elementary partial seizures derived from this area. He
stresses that these sounds' are usually not lateralized, but
occasionally may be referred to the contri,ilateral ear. Perceptual
auditory illusions may also derive from this area.
\XTilliams 7 stresses that the exact topographic localization of a
primary sensory area in the temporal lobe, as found in the
modality of hearing, is usual, and that the surrounding auditory
association areas (Brodmann's areas 42 and 22) are much larger
than the primary area and are responsible for auditory
hallucinatory qualities: these are more complex than just sounds,
and in epilepsy would be complex partial seizures. Midtemporal
lesions may therefore produce undeveloped sounds (e.g.
whistles) or developed sounds (e.g. music), and these may be
associated with a special meaning, relevance or significance.
They are common concomitants in dreamy states; Bingley2
found them in 18% of 29 patients with dreamy states, with
tendency (c = 2,0) to non-dominant localization.
Olfactory hallucinations are classically and a-Iso
statistically usually of unpleasant and disagreeable nature. B
Ervin s stresses the burning or rotting quality that is usual. They
are rare, and may reflect an anteromedial lesion of the
hippocampus. IQ Williams 7 points out that the olfactory cortex
has not actually been lo'calized, and that the previous association
with the anterior hippocampus is being reviewed.
Olfactory hallucinations have been considered by many
authors to be an essential part of the 'uncinate fit' but Bingley2
points out their rarity and that Hughlings Jackson has been
misquoted. Penfield 3 thinks that the rarity of the olfactory
hallucination may put in question the localization of smell in
man; olfactory hallucinations can at best be regarded as a poor
localizer. Feindel and Penfield 9 felt that olfactory hallucinations
more commonly reflect tumours as opposed to atrophic lesions.
Ictal olfactory hallucinations occurred in 2,3% of Lund'sll 733
patients with a supratentorial tumour, of whom three-quarters
had temporal lobe tumours. Fifty per cent of his 34 patients with
'dreamy states' had olfactory hallucinations. Currie ec al. 6found
that 12% (80) of TLE patients have olfactory hallucinations.
It can be concluded that olfactory hallucinations of the
paroxysmal kind suggest temporal lobe origin, and this becomes
much more certain if they are part of a compound complex

903

temporal lobe experience. On their own they are classified as an

elementary partial seizure 8
Gustatory hallucinations. Currie ec al. 6 elicited gustatory
hallucinations in 3% (20 patients) of his TLE series. These also
occurred in 3% of Lund's series of 733 supratentorial tumour
patients; two-thirds derived from temporal lobe tumours, and 10
of the 21 patients involved also had dreamy states. II
The work of Penfield and Faulk l2 suggests that the sensation
of taste can be evoked by stimulation of the insula below the
superior opercula. \XTilliams 7 points out that illusions or
hallucinations of taste are rare in organic lesions. Also worth
noting is the fact that olfactory disturbances are sometimes
experienced as disturbances of taste. 3 Rasmussen B regards
paroxysmal disturbances of taste as being elementary partial
seizures of deep-lying Sylvian origin.
Vestibular symptomatology. Ictal vestibular features arise
from the middle and posterior areas of the superior temporal
gyrus. Anatomically, the primary vestibular area is posterior and
postero-inferior to the auditory cortex. I>The subjective vertigo
is often described as dizziness producing a paroxysmal
impression of 'turning' of self or the environmenr. B This
rotational component is symptomatic of true vertigo. \XTilliams 7
points out that vestibular ictal features can also manifest as
dysequilibrium. Currie ec al. 6 found this to be a feature of 19%
(126) of his TLE patients.

Elementary symptomatology with visceral

sensations or autonomic symptoms
A great variety of autonomic phenomena derive from the periamygdaloid area, but none originate from the lateral temporal
area. 14 Feindel and Penfield 9 found (in their 50 cases) that among
the wide range of phenomena that initiate, accompany and follow
automatisms, flushing or pallor of the face occurred in 22% and
abdominal sensations (including nausea) in 34%. These were, of
course, complex phenomena (part of the dreamy state)
frequently occurring (72%) as compound symptoms.
Visceromotor symptoms may also involve sweating 1S
Alimentary symptoms. Stimulation of the cortex of the
insula produces sensations referring to some part of the
alimentary tract from the mouth to the rectum. Motor responses
such as borborygmi or the desire to defaecate may also occur.>
Penfield and Faulk 12 found that stimulation ofthe anterior insula
produced movement, and posterior insula, sensations; Currie ec
al. 6 in their study found that visceral features occurred in 40%
(264) of TLE subjects. Epigastric sensations, although deriving
from the insula, are looked upon as rather classic features ofTLE
and are variously described. For example, 'sinking feeling' is
Batchelor's description,16 'rising' is Slater and Roth's,1 and
'gripping' that of Penfield. 17

Complex partial seizures of temporal lobe

origin
These form the majority of the epileptic abnormalities occurring
in the temporal lobes and 'complex partial seizures' is now
frequently used as a synonym for TLE.IB In general, they are
associated with defects in consciousness, which may only be
transient, may be associated with partial recall and frequently
have a mixed quality encompassing, for example, cognitive
disturbances (depersonalization, deja vu Qr deja viCll) with
emotional hallucinations.' Partial preservation of awareness is
the special feature of focal epilepsy, particularly the temporal
lobe attack. I \XTilliams stresses that epileptic events are
caricatures of the function of the executive or receptive primary
or association areas.' These result in a 'disintegrated' picture.
This description is particularly applicable to the complex partial
seizure.

904

SA MEDIESE TYDSKRIF 5 DESEMBER 1981

Complex partial seizures with impaired

consciousness only
\1( ilder Penfield 17 describes the 'amnesic seizure' which may
or may not have a motor automatism component. The one
constant characteristic is the subsequent amnesia, even though
the patient may not appear unconscious. The alteration in
consciousness itself may be severe or slight. The amnesia is due
to the loss of the 'capacity of making a permanent record in the
stream of consciousness', the basic defect being an inability to
record present experience. 17
Anatomically, amnesia can be produced by stimulation of the
area around the amygdaloid nucleus which, however, also causes
automatism. 16 As a sequel to this finding, Penfield,3 with Milner,
was able to show that the hippocampal zone played an important
role in the recording of present experience. This could be the
anatomical site of the epileptic absence frequently seen in TLE.
A complex partial seizure with loss of consciousness involves
an 'absence'. These absences are said to be differentiated from
the typical petit mal generalized absence by the concentrated
'fixed stare', 'not blank eyes', which may be indiscriminate at
some poiin, and last from 30 to 60 seconds. 19 These must also be
differentiated from aphasic anacks.
Another feature found either alone or in combination in the
'dreamy state' is what Feindel and Penfield 9 described as
'conscious confusion'. This symptom was described in 32% oftheir
50 cases. Its anatomical localization is also probably in the
anterior mesial temporal area.

Complex partial seizures with cognitive

symptomatology
These can take many forms. For example, a classic symptom is
the 'play-back' in which there is a reliving of stereotyped
auditory or visual memory as part of the ictal experience. I5The
best known research is that ofPenfield and Perot 20 who in more
than 1000 operations elicited visual and auditory experiential
recall by stimulation of areas of the lateral temporal lobes. The
most useful classification of cognitive complex temporal lobe
experiences iS'that of Wilder Penfield 17,21 who called these 'flashbacks' ('play-backs') 'psychical hallucinations' or 'experiential
seizures'. These involve hallucinations of past experience, and
reactivation of a strip of the stream of consciousness.
The 'interpretative seizure' is Penfield's second cognitive
temporal lobe complex component. These involve alterations of
the present and illusions (or false interpretations) of present
perception. Penfield 3 classifies these illusions into auditory
illusions of distance, loudness and tempo, visual illusions of
comparison involving familiar~ty, strangeness and unreality, and
emotional illusions of fear, separation, sorrow and disgust. An
overlap occurs here: interpretative illusions of comparison come
under cognitive dysmnesic disturbances in the form of
familiarity illusions, under which Penfield classifies deja VII;
illusions of emotion fall into Gastaut's~ classification of ictal
emotion; and illusions of auditory and visual kind fall under
Gastaut's 'psychosensory illusions'. Penfield's 'experiential
seizures' are largely synonymous with Gastaut's 'ideational
disturbances' .
During both types of anack the patient continues to be
conscious and he is usually capable of introspection.1 6 Penfield
regards their origin from the same area (lateral and superior
temporal lobe of both hemispheres, excluding primary auditory
and visual areas, but including Wernicke's area) to be significant
for localization of what is now regarded as memory engrams.
Under the heading of djsmnesic disrurbances, Gastaut 4 has
included conscious amnesias, deja vu and deja vecu. Also
included for convenience, although their memory disturbance
components may be disputed by individual authors, are the other
illusions of comparison involving strangeness .and unreality namely jamais vu, depersonalization and derealization.

Williams 7 points our that disturbances of memory are

associated with bilateral anterior temporal lesions, frequently
linked to naturally occurring components such as sight, sound,
smell and emotional tone. The crucial interplay of memory recall
with subjective awareness of time is also stressed by Williams. 7
Gastaur includes forced thinking and dreamy states under
'idearional dislUrbances'. Dreamy scares have been dealt with
earlier and are also discussed later. Forced rhinking involves the
intrusion into consciousness of a thought that is resisted, but
remains or repeats itself ('compulsive thinking'). Reef2 2 points
out that it is usually unpleasant. Forced thinking is, of course,
usually symptomatic of an obsessive-eompulsive neurosis. Its
value in the diagnosis of temporal lobe epilepsy is, therefore, only
nonspecific.

Complex affective symptomatology

\1( illiams 7 regards these symptoms as being ~emotional
hallucinations'; they are usually unpleasant, vivid and unrelated
to the environment,IO and often autochthonous. In 100 epileptics
with these he found 61 to be of anxiety quality, 21 depres~ive, 9
pleasant and 9 unpleasant. He also described feelings of anger
and suspense. The pleasant and unpleasant feelings, he felt,
had posterior temporal origin, while the anxiety had anterior
temporal origins.
Ervin 5 stresses that temporal lobe affect can present as
depression which can take the form of acute despair, with or
without suicidal ideation. On the other hand, pleasurable
feelings can reach the level of elation, or the subject may
experience serenity. The importance of the temporal lobes in
these affective disturbances had been illustrated by the fact that
operations on the amygdala, as well as medial temporal
lobectomy and bilateral temporal lobectomy, induce change.?
Currie el at. 6found emotional disturbances in 19% (130) of their
TLE sample.
Ictal anxiety: the frequent occurrence of 'fear which comes
by itself in the 'uncinate group of seizures' was' noted by
Hughlings Jackson. 23 MacLean24 confirms that this is an
epileptic phenomenon, out of the psychological context and not
due to natural fear of the fit itself. Williams Io found ictal anxiety
in 3% (61 out of 2000) of epileptics, 165 of whom had complex
ictal symptoms (probably of temporal lobe origin). He attributed
this emotional hallucination to the temporal cortex (35 of the 61
had anterior TL foci, and 17 middle TL foci). Lund 11 found that
60% (10 out of 17) of his sample with ictal an.xiety had other
autonomic disorders (e.g. epigastric sensation, flushing,
sweating, palpitations). Williams 10 found this association in 50%
of his ictal anxiety cases. Bingley2 shows a close statistical
correlation between the topographical and physiological
relationship of ictal anxiety and epigastric sensations. According
to Reef,22 fear is the commonest form of seizure with affective
disturbance, and is usually associated with a hallucinatory
experience. Ictal anxiety may also be experienced as panic
symptoms. 15
Aggressive outbursts: The anatomical location of
aggression has important connections in the amygdala and
periamygdaloid areas. 25 This has been supported by EEG
findings. Thus, Williams 26 points to EEGs with predominaiuly
anterior temporal and frontotemporal abnormalities in 2000
subjects with habitual aggression. This finding is of nosological
interest too, as the disturbed behaviour frequently found with
anterior temporal dysfunction is often regarded as part of the
antisocial personality. Despite all this, aggression or rage as an
epileptic phenomenon of significant intensity is generally rare;
Williams IQ and Feindel and Penfield 9 -luote figures of the order
of 0,5 - 1%. It is uncertain if these operationally refer only to
epileptic furor states, which would be one extreme.
Reef22describes 'seizures with antisocial behaviour', involving
moodiness, irritability, anger or other behaviour disturbances
usually lasting only a few minutes but possibly persisting for

SA MEDICAL JOURNAL 5 DECEMBER 1981

hours. During this time the temporal EEG will become more
abnormal. Children are particularly singled out, but Livingston
and Pauli 27 stress that many of these children outgrow these
symptoms and great caution must be exercised prior to a TLE
diagnosis on the basis of temper tantrums, rages, and grossly
abnormal motor performance. Bingley2 found anger in 17% of
cases, particularly prevalent in the non-tumour group with
bilateral temporal foci (27%). According to Rompel l9
unprovoked temper anacks are 'probably the most common sign'
of TLE. It is difficult t9 reconcile this with the formal research
done, but possibly the psychiatric population ofTLE patients is
vastly different to the neurological population. A more
conventional explanation would be that extreme caution should
be exerted before a diagnosis of TLE on the basis of 'temper
anacks' is made, and that only uncontrolled, unprecipitated,
relatively undirected anacks, often involving violence or injury
to self, and with some degree of defect of consciousness or
amnesia for part of the episode can now be regarded with any
degree of specificity, unless they form part of a localizing
compound complex seizure panem. Explosions of irritability
may also occur as a post-ictal phenomenon, associated with
aggress"iveness which may be severe.'

Complex partial seizures with psychosensory

symptomatology
Depersonalization and derealization phenomena are
frequently associated with illusions and hallucinations,
according to Williams,7 and are just two striking manifestations
that may compound psychosensory temporal lobe
symptomatology. They may occur in combination with many
other temporal lobe symptoms, as confirmed repeatedly in the
literature (Feindel and Penfield,9 Currie er al., 6 Bingley2). These
illusory and hallucinatory experiences may also be associated
with a defect in consciousness, producing a 'dreamy state'. They
involve complex seizures because of their experiential/
interpretative distorting or disintegrating quality. Their actual
content may be protean and widely varied, but their frequent
stereotyping and constancy within individual patients is
noteworthy.'
Psychosensory illusions may be visual or auditory.
Penfield 3 regards 'visual illusions of alteration' occurring as
epileptic .discharges, as being of predominantly non-dominant
(for handedness, not speech) temporal hemisphere origin (10 out
of 11 cases). He also elicited these features by only stimulating
the minor hemisphere. If these results are generalized, it
suggests that visual psychosensory illusions may have excellent
lateralizing value. Penfield's subcategorization of visual illusions
is extensive. He notes firstly, alterations in clearness (Feindel
and Penfield 9 noted visual blurring in 16% oftheirsample of 50).
The second alteration may be one of distance and nearness
(seeing objects as too far or too near) and this is closely allied with
distortions in size (macropsia or micropsia) which is not
specifically stressed by Penfield. 3 Distortions of shape may also
occur - metamorphosia falls under this category. Objects may
appear to be racing, or moving too slowly, or stationary when
moving, or vice versa (illusions of speed of movement). Finally,
distortions in angulation/erectness perception may occur.
Penfield 3 stresses that interpretation of position, shape, clearness
and speed of things depends particularly on the function of the
non-dominant hemisphere. Today, the spatial function of the
minor hemisphere is well recognized,26 so that logically these
visual illusions are minor hemispheric.
In contrast to visual illusions, Mullan and Penfield 29 did not
find any difference in frequency of alldilOry illusions on
stimulation of either temporal lobe. It appears, therefore, that
auditory illusions are of no localizing significance (20 cases).
There are three major kinds of auditory alterations of perception
of the present. 3 Illusions of loudness may occur, but illusions of

905

distance or of tempo have been evoked by stimulation of the first

temporal convolution or clo e to it. Mullan and Penfield 29
showed a close anatomical relationship of auditory illusions to
illusional emotions, which occurred slightly inferior to the
auditory illusions and were also bilateral phenomena. The main
emotion stressed was fear, but illusionalloneliness, separation,
sorrow and disgust were also described. 3
Psychosensory hallucinations: Wilder Penfield's3
'reactivation of a strip of the record of the stream of
consciousness' may involve the special senses, in which case a
psychosensory hallucination of temporal lobe origin occurs.
These are complex partial phenomena. \XI'hen they occur
paroxysmally, they are complex partial seizures. As these
psychosensory hallucinations theoretically involve a reactivation
of something previously in memory storage, they hould differ
markedly from the hallucinations found in the functional
psychoses. However, a variety of other compound complex
phenomena of cognitive, affective and other psychosensory kind
may be associated with the reactivation of the stream. This may
produce a picture very similar to an acute schizophrenic or manic
episode.
Penfield 3 points out that the epileptic is having the double
experience of a re-creation of the past with the consciousness of
the present time during the experiential hallucination. As a
specific area may theoretically be involved in firing during a
seizure, the epileptic will have the same psychosensory
hallucination repeatedly. It may also not have the specific
relevance, special meaning, auditory quality, and self-reference
so frequently found in schizophrenia.
The psychosensory hallucination deriving from the temporal
lobes can involve any of the sensory modalities or combination
thereof, namely visual, a uditory, olfactory, gustatory, vestibular,
tactile or the 'indescribable' hallucinations described by
Williams. 7
The complexity of these hallucinations depends upon the
specific area of firing. Williams'7 conceptualization of increasing
qualitative complexity of each experience as the area of firing
moves away from the primary sensory to the secondary, tertiary
and more distant association areas, is again emphasized.
Williams 7 describes TLE hallllcinacions of indescnbable qllaliry
which may originate in the medial temporal cortex; their
indescribable quality may be associated with pathways not
usually brought into conscious awareness.

Psychomotor seizures and the 'dreamy state'

These have already been discussed. The major feature of the
'dreamy state' of Jackson 30 is 'whole or partial withdrawal from
the present'. This involves a defect of consciousness of the
environment, but not of consciousness ofself. The 'dreamy state'
emphasize
the subjective mental phenomena. The
'psychomotor seizure' emphasizes behavioural phenomena.
Technically, the two are almost identical. 2 As both terms are
used to refer to a wide variety of combinations of perceptual and
motor disturbance, these symptoms are best discussed
individually, with the knowledge that compound partial seizures
(usually complex) in many combinations make up the mental and
behavioural components of the dreamy state or psychomotor
seizure.
Automatisms are 'complex partial seizures with
psychomotor symptomatology'.4 The automatism involves the
motor component of the psychomotor seizure. Its usual origin is
from the mesial temporal area,14 never the lateral temporal
area. 31 It may, however, have orbitofrontal, subcortical and other
extra-temporal lobe originsY Further evidence as to its origin
relates to the success of treatment of automatism by operative
ablation of the amygdala and surrounding tissue, and production
via stimulation of this area 9

906

SA MEDIESE TYDSKRIF 5 DESEMBER 1981

Automatisms have a stereotyped fixity. They may be

behavioural, involving masticarory/salivatory, buttoning/
unbuttoning or other simple, repetitive movementS, and seldom
last more than a few minutes, often far less.
Alternatively, Ervin 33 stresses the rich flow of associations
achieved under hypnosis and the induction of amomatisms by
this method. He points om that in the same way that normal
dreams have symbolic significances so too do complex
automatisms. These may last hours or, rarely, days. A hazy or
distorted memory may be epileptic. Epileptic automatisms
involve programmed motor activity set off by the attack 22 and
interference with this will cause resistance which may be
dangerous. The attack often begins with a 'dreamy state' of some
kind. Automatisms may take the form of fugue states, speech
automatisms or furor, and involve very complicated acts with
normal or occasionally dyssocial behaviour. They may be
differentiated from hysterical dissociation by their abruptness,
the absence of precipitating factors, and the complete amnesia
(Livings ton and PauIi 27 ). Ervin 5 stresses the absence of
secondary gain, and the fact that even under hypnosis the
amnesia persists. Nemiah 34 points out that clinical dissociative
features with other evidence of TLE should be suspected to be
organic.
The electroclinical correlate may be forced normalization or
an unchanged EEG, in which case depth electrography may
clarify the picture. Depth electrography (using sphenoidal
electrodes) with chemical activation of the patient's habitual
seizure recorded is suggested in cases of difficult diagnosis by
Remick and \X ada. 18 Ervin 5 differentiates the post-ictal twilight
state, which commonly presents with automatic behaviour
associated with a confusion, lasting often days, and with
symmetrical, slow, non-paroxysmal EEG activity.
Bingley 2 points out that in ictal speech aUlOmacisms, the patient
utters a mixture of words and sentences which may be
linguistically correct but bear no appropriate relation to the
present situation; this is less closely related to the dominant lobe
than ictal aphasia and more closely related to other forms of
automatism. Thirty-three per cent of his sample had speech
automatism.
Rare psychomotor symptoms include argasmalepsy, in
which the patient may experience orgasm as an aura, or in
association with other symptoms e.g. laughter or weeping.
Anatomically this may be associated with the limbic part of the
temporal lobe, but may also occur with midline di-encephalic
lesions. 35 Gelascic epilepsy involves paroxysms of laughing or
giggling, usually associated with pleasure. Quiricarian (ar
dacracyscic) epilepsy invovles crying episodes. Cursive epilepsy
involves actual running in paroxysms.
Gelastic, dacrocystic and .cursive epilepsy are all highly
unusual but well described epiIeptic symptoms. These are
usually associated with amnesia and often involve the complex
juxtaposition of temporal lobe symptoms of the visceral,
expressive and affective kind. They may derive from the
amygdaloid or hippocampal areas of the temporal lobe, but may
also have origin in the posterior hypothalmus, basal ganglia and
third ventricle. 36

Partial temporal lobe seizures, secondarily

generalized
Bingley 2 found that 83% of his sample of 90
neurological/neurosurgical patients had tonic-clonic seizures
secondary to a primary temporal lobe focus. There was no
significant difference between tumour and non-tumour cases, or
in laterality of the lesion. This figure corresponds closely with
others quoted for associated tonic-clonic seizures, the exact
sample derivations of which are unknown: Gibbs 37 and
Anderson and Trethowan 15 both indicate that in 80% of cases
TLE is associated with grand mal seizures. The latter stress that

almost all cases of aura followed by tonic-clonic seizure originate

from the temporal lobe. It is very possible that the figures for
generalization of seizures of temporal lobe origin are far lower in
psychiatric populations. In any event, the presence of tonicclonic seizure with auras oftemporallobe origin can be regarded
as making the diagnosis of TLE almost definitive.

Post-ictal symptomatology
The presence of typical post-ictal features may alert the
clinician to a diagnosis of epilepsy, although these may reflect
epilepsy of generalized, partial or unclassifiable origins. Postictal features are common, and unless at least one post-ictal
feature typically occurs with seizures which generalize
themselves (particularly tonic-clonic seizures), an equivocal
diagnosis of epilepsy should be seriously reviewed.
The most common features are listed by Slarer and .Roth I as:
confusian: clouded consciousness with disorientation for time,
place and even person, often with rambling speech or disjointed
behaviour, fidgeting or restlessness; pasc-epilepcic amarnacisms
that may involve well-controlled behaviour or purposeless
movements; sleep with fatigue; headache, often very severe, after
waking from a post-ictal sleep or separately. Nausea with or
without vomiting is also common.

Aetiology of temporal lobe epilepsy

It is important to summarize the most common causes ofTLE,
as the presence of a recognized aetiological factor in the history
may alert the clinician to the diagnosis.
In the large London Hospital sample (about 660) of Currie el
al. 6 two-thirds of TLEs were of unknown aetiolgy. Tumours
were found in 9,5%, head injury was associated with 7%, as was
birth trauma. Infantile seizures had occurred in 5%,and II %had
a family history ofepilepsy. Overall, 25% of cases were thought to
have a definite cause. This study found no EEG temporal lobe
focus in 8%, and in only 6% was the 'diagnosis purely
electrographic. Of the EEG abnormalities, 52% were left-sided,
29% right-sided and 19% were bilateral. Routine activation
procedures, often with sleep and sometimes with sphenoidal
leads, were used.
Penfield and Jasper 38 explained the frequency of mesial
temporal lesions as due to 'incisural sclerosis' which results from
the bitemporal scarring associated with compression of the
anterior choroidafartery due to herniation of the deepest portion
of the first temporal convolution, uncus and hippocampus
during birth trauma. The alternative explanation relates to
infantile hyperpyrexial convulsions which may affect this area,
which is relatively primitive and highly sensitive to anoxia. 39
It is particularly important to elicit any history of birth
trauma, infantile febrile convulsions, head injuries and
constitutional predisposition to epilepsy. Other possible causes
such as the presence of tumour, infection and p.r:enatal and
postnatal problems, should also be noted.

REFERENCES
!. Slater, E. and Roth, M. (1969): Clinical Psychiacry, 3rd ed. London: Bailliere
Tindall.
Bingley, T. (1958): Acta psychiat, scand., 33, suppl 120, p.!.
Penfield, W. (1958): Res. Publ. Ass. nerv. men!. Dis., 36, 210.
Gastaut, H. (1970): Epilepsia, 11, 102.
Ervin, F. R. in Freedman, A. M., Kaplan, H. 1. and Sadock, B. J., eds (1976):
Comprehensive Texcbook of Psychiarry, vol. 1, pp. 1138 - 1157. Baltimore:
Williams & Wilkins.
6. Currie, S., Heathfield, K. W., Henson, R. A. ec al. (1971): Brain, 94, 173.
7. Williams, D. in Vinken, P. J. and Bruyn, G. W., eds (1977): Handbook of
Clinical Neurology, vol. 2, pp. 700 - 724. Amsterdam: North-Holland.
8. Rasmussen, T. ill Vinken, P. J. and Bruyn, G. W., eds (1977): Op cil.', vol. 15,
,
pp. 74 - 86.
9. Feindd, W. and Penfield, W. (1954): Arch. Neurol. Psychiar. (Chic.), 72, 605.

2.
3.
4.
5.

SA MEDICAL JOURNAL 5 DECEMBER 1981

10. Williams, D. (1956): Brain, 79, 29.
11. Lund, M. (1952): Epilepsy in Associatian with Imracranial TumiJllr.
Copenhagen: Munksgaard.
12. Penfield, W. and Faulk, lVL E. (1955): Brain, 78,445.
13. Carmichael, E. A., Dix, M. R. and Hallpike, C. S. (1954): Ibid., 77, 345.
14. Feindel, W. in Vinken, P. J. and Bruyn, G. W., eds (1977): Op cit', p. 87.
15. Anderson, E. W. and Trerhowan, W. H. (1973): Psychiatry, 3rd ed. London:
Bailliere Tindall.
16. Barchelor, I. R. C. (1969): Henderson and Gillespie's Textbook 0/ Psychiatry,
IOm ed. London: Oxford Universiry Press.
17. Penfield, W. (1955): J. memo Sci., 424, 451.
18. Remick, R. A. and Wada, J. A. (1979): Amer. J. Psychia!., 136, 320.
19. R"mpel, H. (1966): Med. Proc., 12, 175.
20. Penfield, W. and Perot, P. (1963): Brain, 86, 595.
21. Penfield, W. (1954): Bri!. J. Surg., 41,337.
22. Reef, H. E. (1970): Leech, 40, 63.
23. Jackson, H. J. and Srewan, P. (1899): Brain, 22, 534.
24. MacLean, P. D. (1952): Elecrroenceph. clin. Neurophysiol., 4, 407.
25. Anand, B. K. and Brobeck, J. R. (1951): Yale J. BioI. Med., 24, 123.
26. Williams, D. (1969): Brain, 92, 503.
27. Livingsron, S. and Pauli, L L in Freedman, A. M., Kaplan, H. I. and Sadock,

907

B. J., eds (1976): Up cit." vo! 2, p. 2060;

28. Omsrein, R. E. (1975): The Psychology o/Consciollsness, pp. 65 - 9. New York:
Pelican.
29. Mullan, S. and Penfield, W. (1959): Arch. ~eurol. Psychia!. (Chic.), 81, 269.
30. Jackson, H. J. (1888): Brain, 11, 179.
31. Schneider, R. c., Crosby, E. C. and Calhoun, H. D. in Kahn, E. A., Crosby, E.
c., Schneider, R. C. et aI., eds (1969): Correlative Neurosurgery, pp. 279 - 358.
Springfield, Ill.: Charles C. Thomas.
32. Pemy, J. K. and Dreifuss, F. E. (1969): .-\rch. ~eurol. (Chic.), 21,142.
33. Ervin, F. R., Epsrein, A. W. and King, H. E. (1955): Arch. Neurol. Psychia!.
(Chic.), 74, 488.
34. Nemiah, J. C. in Freedman, A. M., Kaplan, H. I. and Sadock, B. J., eds
(1976): Op ci!. " pp. 1230 - 1231.
35. Shukla, G. D., Srivaslava, O. N. and Kaliyar, B. C. (1979): Bri!. J. Psychia!.,
134, 288.
36. Leading article (1977): Bri!. med. J., I, 127.
37. Gibbs, F. A. (1951): J. nerv. memo Dis., 113, 522.
38. Penfield, W. and Jasper, H. H. (1954): Epilepsy <lnd the Fllnctioll<ll rJn<llOlt,r 0/
[he Human Brain. Boston: Linle, Brown.
39. Earle, K., Baldwin, M. and Penfield, W. (1953): Arch. Neurol. Psychia!.
(Chic.), 69, 27.

Is -deja vu a symptom of temporal lobe

epilepsy?
V.M. NEPPE
Summary
The definition and conceptualization of the deja vu
phenomenon are interpreted in various ways. The
common occurrence of deja vu in the general
population stresses the need forthe development of
specific qualitative features which will be valuable in
the diagnosis of temporal lobe epilepsy.
s. Air. med. J .. &0,907 (1981).

Definitions
Exact defInitions of the deja. vu phenomena differ from that of
Warren,1 who refers to an 'illusion of recognition', to others that
do not require the recognition concept at all, such as Barton's2
'an abnormal feeling offamiliarity'. Some defInitions do not even
stress the erroneousness of the experience, and that ofRycroft 3 is
probably the best example: 'The subjective sense that a present

novel experience has been gone through subjectively'. Thus deja

vu is conceptualized differently, and at times contradictorily.
One important theoretical contradiction involves part-whole
recognition; Silverman4 stresses how deja VII illustrates 'redintegration in which the same stimulus or stimuli has occurred
at another time (and) elicits a pattern of remembering'. This
implies that a 'parr tends to reinstate the whole'> and is
equivalent to what Banister and Zangwill 6 call a 'restricted
paramnesia', being specifically differentiated from deja VII which
is 'the illusion of reliving a situation in its entirety'.
Deja VII strictly implies an 'already seen experience', but in a
broader sense can imply any feeling of familiarity with the past.
In this study the concept of deja VII is used in its broader sense. It
therefore includes many other experiences of the past: deja
emendll (already heard), deja eprollve (already experienced), deja
fail (already done), deja pense (already thought), deja 71011111
(already desired), deja raCOnle (already recounted), deja semi
(already felt), and deja vecu (already lived).
Deja vecu: Strauss 7 conceived of aspecial kind of deja vu with
a special qualitative component going far beyond familiarity
through to the impression of reliving the same situation. It is
possible, bur apparently unproven, that this specific kind of dija
VII occurs signifIcantly more often in temporal lobe epilepsy
(TLE).7

Department of Psychiatry, University of the Witwatersrand,

Johannesburg

v. M. N EPPE,

M.B. B.CH., D.P..\1., B.A., F.F.PSYCH.(S.A.), "DIED. (PSYCH.),

Categorization

PH.D. (MED.)
Dare received: 19 December 1980.

The deja vu phenomenon can be categorized as a paramnesia, as

an ego-state disorder, or as a perceptual disturbance. All these