AWA: Academic Writing at Auckland

Retrograde amnesia is a phenomenon which has brought about much debate including the development of two partially opposing theories, the Standard Consolidation Model and Multiple Trace Theory. Each proposes different explanations for the role the hippocampus plays in the storage and recollection of remote memories and predicts the effects damage to this medial temporal lobe area will have. Over the years, extensive research, driven by notable individuals such as Squire and Moscovitch, has been conducted with a focus on declarative memory. Results obtained have been interpreted to support each respective theory. Whilst each is convincing, the refinement of methods used for testing (developed as a result of the conflict between the two theories), means that overall, observations seem to support the Multiple Trace Theory in being the more plausible of the two. However, despite all this, it does not yet seem possible to clearly establish the role of the hippocampus considering the unclean nature of lesions in case studies as well as the differing views of the theories regarding its contribution in relation to the medial temporal lobe and its other component structures.  

According to the Standard Consolidation Model, the hippocampus plays a time-limited role in the storage and retrieval of declarative memory until this function is eventually taken over by the neocortex where it remains permanently. Squire and Alvarez (1995) argue that this transfer of memory (both semantic and episodic) occurs via a time-dependent consolidation process during which the repeated reactivation of representations results in the formation of interconnections between “separate regions of neocortex” (Alvarez & Squire, 1994). This occurs gradually as “neocortical synapses change slowly” (Squire & Alvarez, 1995) and eventually relieves any dependence on the hippocampal system.

One of the studies used to support this view of a time-limited hippocampal role was carried out by Squire (1990) and involved the bilateral removal of the hippocampal formation in monkeys. Training to discriminate a selection of 100 objects occurred 16, 12, 8, 4 and 2 weeks prior to surgery. Following this, the results of a single-choice trial were compared with those of normal monkeys who had undergone the same training but not surgery. It was found that the normal monkeys performed better overall as one would expect. They also “exhibited forgetting” (Zola-Morgan & Squire, 1990) - being more able to recall recently learned items as compared to those learnt at 12 and 16 weeks prior. Contrastingly, it appeared that there existed a temporally graded retrograde amnesia in the operated monkeys, as reflected in their scores which were significantly higher for objects learned 12 weeks prior (72.3% correct) than for those learned 2 weeks prior (62.3%). As such, this study appears to support the proposition that remote memory is relatively impervious to hippocampal damage - a result, the SCM says, of consolidation which has allowed it to become independent in the neocortex. It is interesting to note that the H⁺ lesions the monkeys received included the hippocampus proper, dentate gyrus, subicular complex, posterior entorhinal cortex and most of the parahippocampal cortex. This study will be further discussed later.

The Multiple Trace Theory, on the other hand, proposes that the hippocampal system always remains involved in memory retention and recollection. This, as also discussed later, is particularly so for episodic autobiographical memory. According to Viskontas, McAndrews and Moscovitch (2000), “the hippocampal complex rapidly binds novel information and experience into a coherent memory trace” and from then on, each time this old memory is retrieved, a new, unique trace is established (Moscovitch, et al., 2005), ultimately creating a network of multiple hippocampal-neocortical traces.

There are two underlying issues which are crucial in separating the two theories. The first of these is whether or not there exists a difference between remote episodic and semantic memory and their respective relationships with the hippocampus. In their definition of episodic memory, Moscovitch and his colleagues (2005) use two key phrases to distinguish it from semantic memory – “spatial-temporal context” and “re-experiencing”. They believe that episodic memories are those “for particular, autobiographical episodes that have a distinct spatio-temporal context and involve a detailed re-experiencing of the initial event” (Moscovitch, et al., 2005). That is, that the individual is somewhat able to travel back in time during recall. Semantic memory, on the other hand, is general knowledge which is “free of context” (Steinvorth, Levine, & Corkin, 2005). It includes knowledge of the world as well as information related to oneself and it, together with episodic elements, constitutes autobiographical memory. The Standard Consolidation Model predicts that both types of memory are equally affected by damage to the medial temporal lobe and result in a temporally graded retrograde amnesia (Squire, Stark, & Clark, 2004). The Multiple Trace Theory does not dispute the gradual independence of semantic memory and thus corresponds with the Standard Consolidation Model in this area. However, with regards to vivid episodic memory as defined above, it expects that damage to the hippocampus will result in an ungraded retrograde amnesia, reflecting its continued involvement in episodic retention and recall (Steinvorth, Levine, & Corkin, 2005).

In their report published in 2005, Steinvorth and her colleagues conclude that results following tests on two amnesic patients with bilateral MTL lesions, H.M. and W.R., are supportive of MTT views as described above. Using the Autobiographical Interview method, it was found that W.R.’s scores for the number of external autobiographical details recalled were within the range of the eight controls, whereas she exhibited a deficit for internal details (Steinvorth, Levine, & Corkin, 2005). H.M. performed poorly on the New Vocabulary Test and Facts of Public Events Test (both testing for semantic memory) – corresponding with both the SCM and MTT, and also displayed a “severe autobiographical deficit, being only capable of giving “script-like descriptions” of events such as “falling in love” and “taking a test at school” (Steinvorth, Levine, & Corkin, 2005). Collectively, the results add weight to the view that episodic and semantic elements of autobiographical memory are differentially affected by damage (Moscovitch, et al., 2005).

This view is additionally supported by similar results from experiments such as those carried out by Viskontas, McAndrews and Moscovitch (2000) in which twenty-five patients suffering from temporal lobe epilepsy (TLE) were tested alongside twenty-two controls matched for age and education using the Autobiographical Memory Interview. TLE patients easily recalled personal semantic memories but displayed a “profound impairment”, as with H.M. and W.R., in autobiographical episodic recall and along with this, no temporal gradient (Viskontas, McAndrews, & Moscovitch, 2000). However, it must be noted that there existed a range of anatomical damage amongst patients and most of the exact details of these were left vague.

This leads into the second and pivotal issue – the existence of large discrepancies between studies on precisely which structures of the medial temporal lobe are responsible for observations. This is coupled with the fact that the lesions suffered in case studies vary from individual to individual, making it difficult to pinpoint the precise roles of various structures. Those in support of the MTT, most notably Nadel and Moscovitch (2005), put forward that for any episodic memory, recent or remote, the entire medial temporal lobe region (that is, the hippocampus proper, dentate gyrus, subiculum, entorhinal cortex, perirhinal cortex and parahippocampal cortex), is always necessary (Moscovitch, et al., 2005). Their proposition that the “multi-faceted nature of autobiographical memory” (Moscovitch, et al., 2005) being a reflection of the differing roles each structure plays, is reasonable. It is, however, difficult to develop and support this suggestion at this time as studies specifically targeting these various structures would need to be conducted.

Should Moscovitch and his colleagues (2005) be correct, what should be observed is that with any large areas of damage to the medial temporal lobe, an ungraded retrograde amnesia is developed for vivid episodic memory. Their point of view also opens another possible interpretation to the temporally graded retrograde amnesia some have reported – that if the cause was a small, limited lesion to the medial temporal lobe, remote memories (which have had a longer period of time over which multiple traces could have been formed) are less affected as their traces are more widely spread throughout the region (Moscovitch, et al., 2005). However, this is not the case for those in favour of the Standard Consolidation Model. In their report, Kapur and Brooks (1999) differentiate between the hippocampus and what they describe as the “neocortical areas of the temporal lobe” – the entorhinal, perirhinal and parahippocampal cortices (Kapur & Brooks, 1999). They suggest that these structures function as the neocortex in which memories are stored following consolidation. This same implication is made by Bayley, Squire and their colleagues (Bayley, Gold, Hopkins, & Squire, 2005) who compared results between patients with bilateral lesions restricted to the hippocampus, dentate gyrus and subicular complex and those with large medial temporal lobe lesions.

Additional to the studies already mentioned are further neuropsychological and neuroimaging studies used to support each theory. The aforementioned animal study in support of the SCM (Zola-Morgan & Squire, 1990) is an example of neuropsychological research. Though not completely irrelevant, the weight of their results seems to diminish with the consideration of the fact that the H⁺ lesion administered included structures Squire and other colleagues would later categorise as neocortex (Kapur & Brooks, 1999; Bayley, Gold, Hopkins, & Squire, 2005) . Furthermore, its being an animal study makes it impossible for communication with the patients to occur and thus also impossible to assess whether or not it was semantic or episodic memory that was relied on. The ability to measure the vividness of any episodic elements is also eliminated. Studies not clearly distinguishing between the two memory types is a trend seen in much of the earlier work carried out prior to the introduction of the MTT. This is also seen in Squire and Ablarez’ s study (1995) which cites the examples of four patients – R.B., G.D., L.M. and W.G., all of whom suffered bilateral damage to their hippocampal regions and displayed a temporally graded retrograde amnesia.

In their 2005 study, Bayley and colleagues carried out experiments on five amnesic patients, with a focus on the recall of remote autobiographical memories. The patients were separated into MTL and MTL+ groups depending on the locus of their lesions as shown in their MRI scans. In the tests involving the general recall of autobiographical memories, MTL patients showed no significant difference as compared to controls whilst MTL+ patients performed significantly poorer (Bayley, Gold, Hopkins, & Squire, 2005). The methods described were thorough and support the SCM quite comfortably. However, when rating the vividness and perspective (whether they were re-experienced in first or third person) of the memories, only one of the three MTL+ patients participated and thus, though it appears that the MTT suggestion (that episodic memories high in internal details permanently rely on the hippocampus) is disproved, the significance of the results is compromised.

A key point to note is that whilst improvements have been made to the heavily relied on Autobiographical Memory Interview (used in most of the aforementioned studies), it still involves the assessment of the level of re-experiencing. However, as Gilboa (2004) points out, there currently does not exist a direct measure of this part of memory and thus correlations (such as vividness) are used instead (Moscovitch, et al., 2005). Since this is rather standardised across studies, any major inaccuracies would be accounted for through systematic error. Nevertheless, though difficult, this appears to be an area in need of further development.

Alongside neuropsychological studies are neuroimaging studies. They, as opposed to lesion studies (which observe the effects of an excision or damage), allow for the activation of undamaged structures to be considered instead. According to the SCM, the recall of recent memories should cause greater amounts of hippocampal activations than for remote and the opposite should occur for the neocortex (Moscovitch, et al., 2005). However, for the MTT, activation of the hippocampus should be equal regardless of the age of the memory, “as long as they are vivid or detailed” (Moscovitch, et al., 2005). In earlier work, patients would first undergo a pre-scan interview and following this, memories would be probed for during scanning. Though this method, as carried out by Maguire (2001), yielded results supporting the MTT, it has been criticised for its use of pre-scan interviews and self-selected memories, which open the possibility that the hippocampal activation observed may be a result of the re-encoding of memories as opposed to an initial retrieval (Moscovitch, et al., 2005).

As such, Gilboa’s study (2004) is significant in its use of personal photos (pre-selected by family members) the patient had not seen recently. Patients were asked to re-experience the event in the photo and the extent of activation was compared with a baseline condition. Overall, results showed no difference for remote and recent photos as well as a “greater distribution along the hippocampus in remote memories” – both of which refute SCM predictions in support of the MTT. (All this whilst noting that the exact MTL structures activated is not immediately clear). Gilboa (2004) goes on to mention that previous neuroimaging studies which have given evidence “consistent with reports of temporally graded retrograde amnesia following damage to MTL and related structures” (Gilboa, Winocur, Grady, Hevenor, & Moscovitch, 2004) are confounded by the fact that recent memories tend to be more vivid than remote ones and as such, age becomes a variable for which “attempts to control have been inadequate” (Gilboa, Winocur, Grady, Hevenor, & Moscovitch, 2004).

On the whole, evidence from both neuropsychological and neuroimaging studies seem to weigh more heavily towards the MTT. This is especially so in terms of its distinction between episodic and semantic memory as well as its clear definition of the medial temporal lobe as its focus. However, what is not so clear across both theories is the exact role the hippocampus plays. This is a result of the individual differences existing between patients - the locations of their brain damage (which is usually never a clean cut of individual structures) as well as other confounding factors (such as patient background and age). From the research conducted, it appears plausible to believe that the medial temporal lobe structures function as a unit as the MTT suggests. Yet, this is confounded again by the fact that it still does not allow room for the isolation of hippocampal function and that it is still relatively unclear what contribution each MTL structure makes and the extent of their effect on remote memory as a whole.

Works Cited

Alvarez, P., & Squire, L. R. (1994). Memory consolidation and the medial temporal lobe: A simple network model. Proc. Natl. Acad. Sci. USA, 7041-7045.

Bayley, P. J., Gold, J. J., Hopkins, R. O., & Squire, L. R. (2005). The Neuroanatomy of Remote Memory. Neuron , 799-810.

Gilboa, A., Winocur, G., Grady, C., Hevenor, S., & Moscovitch, M. (2004). Remembering Our Past: Functional Neuroanatomy of Recollection of Recent and Very Remote Personal Events. Cerebral Cortex , 1214-1225.

Kapur, N., & Brooks, D. J. (1999). Temporally-Specific Retrograde Amnesia in Two Cases of Discrete Bilateral Hippocampal Pathology. Hippocampus , 247-254.

Maguire, E. (2001). Neuroimaging studies of autobigraphical event memory. Philosophical Transactions of the Royal Society of London: Biological Sciences , 1441-1451.

Moscovitch, M., Rosenbaum, R. S., Gilboa, A., Addis, D. R., Westmacott, R., Grady, C., et al. (2005). Functional neuroanatomy of remote episodic, semantic and spatial memory: a unified account based on multiple trace theory. J. Anat. , 35-66.

Moscovitch, M., Westmacott, R., Gilboa, A., Addis, D. R., Rosenbaum, R. S., Viskontas, I., et al. (2005). Hippocampal Complex Contribution to Retention and Retrieval of Recent and Remote Episodic and Semantic Memories: Evidence from Behavioral and Neuroimaging Studies of Healthy and Brain-Damaged People. In N. Ohta, C. M. MacLeod, & B. Uttl, Dynamic Cognitive Processes (pp. 333-380). Springer.

Squire, L. R., & Alvarez, P. (1995). Retrograde amnesia and memory consolidation: a neurobiological perspective. Current Opinion in Neurobiology , 169-177.

Squire, L., Stark, C., & Clark, R. (2004). The medial temporal lobe. Annu. Rev. Neurosci. , 279-306.

Steinvorth, S., Levine, B., & Corkin, S. (2005). Medial temporal lobe structures are needed to re-experience remote autobiographical memories: evidence from H.M. and W.R. Neuropsychologia , 479-496.

Viskontas, I. V., McAndrews, M. P., & Moscovitch, M. (2000). Remote Episodic Memory Deficits in Patients with Unilateral Temporal Lobe Epilepsy and Excisions. The Journal of Neuroscience , 5853-5857.

Zola-Morgan, S. M., & Squire, L. R. (1990). The Primate Hippocampal Formation: Evidence for a Time-Limited Role in Memory Storage. Science , 288-290.