A distinction is made between structural and semantic knowledge, focusing on the possible influences of the latter. To the extent that early vision may be influenced by object-identity, it would seem necessary to evoke compiled transducers to explain such an influence. Compiled transducers may furnish a way in which vision can be and is penetrated.

The target article claimed that although visual apprehension involves all of general cognition, a significant component of vision (referred to as early vision) works independently of cognition and yet is able to provide a surprisingly high level interpretation of visual inputs, roughly up to identifying general shape-classes. The commentators were largely sympathetic, but frequently disagreed on how to draw the boundary, on exactly what early vision delivers, on the role that attention plays, and on how to interpret the neurophysiological data showing top-down effects. A significant number simply asserted that they were not willing to accept any distinction between vision and cognition, and a surprising number even felt that we could never tell for sure, so why bother? Among the topics covered was the relation of cognition and consciousness, the relation of early vision to other modules such as face recognition and language, and the role of natural constraints.

By utilizing new information from both clinical and experimental (lesion, electrophysiological, and gene-activation) studies with animals, the anatomy underlying anterograde amnesia has been reformulated. The distinction between temporal lobe and diencephalic amnesia is of limited value in that a common feature of anterograde amnesia is damage to part of an “extended hippocampal system” comprising the hippocampus, the fornix, the mamillary bodies, and the anterior thalamic nuclei. This view, which can be traced back to Delay and Brion (1969), differs from other recent models in placing critical importance on the efferents from the hippocampus via the fornix to the diencephalon. These are necessary for the encoding and, hence, the effective subsequent recall of episodic memory. An additional feature of this hippocampal–anterior thalamic axis is the presence of projections back from the diencephalon to the temporal cortex and hippocampus that also support episodic memory. In contrast, this hippocampal system is not required for tests of item recognition that primarily tax familiarity judgements. Familiarity judgements reflect an independent process that depends on a distinct system involving the perirhinal cortex of the temporal lobe and the medial dorsal nucleus of the thalamus. In the large majority of amnesic cases both the hippocampal–anterior thalamic and the perirhinal–medial dorsal thalamic systems are compromised, leading to severe deficits in both recall and recognition.

Aggleton & Brown argue that the function of the hippocampus and perirhinal cortex can be dissociated along a spatial/nonspatial dimension. They further suggest that this division corresponds to a distinction between episodic and recognition memory. An analysis of the data, however, fails to support the underlying dissociation.

The effort to identify the neural substrate of episodic recall, though ambitious, lacks experimental support. By considering the data on c-fos activation by novel and familiar stimuli in recognition studies, we illustrate how inadequate experimental designs permit alternative interpretations. We stress that interpretation of c-fos expression changes should be supported by adequate recognition tests.

In general, we endorse Aggleton & Brown's thesis that the neuroanatomy of amnesia comprises two functionally distinct systems, but we are disappointed in the lack of detail regarding the critical functional contribution of the hippocampus. We also take issue with the characterization of the cortical areas surrounding the hippocampus, particularly the decreased emphasis on the cortical input to the hippocampus.

Three questions arising from Aggleton & Brown's target article are addressed. (1) Is there any benefit to considering the effects of partial lesions of the anterior thalamic nuclei (AT)? (2) Do the AT have a separate role in the proposed extended hippocampal system? (3) Should perirhinal cortex function be restricted to familiarity judgements?

Perirhinal cortex contributes to judgements about stimulus familiarity, but its role is far greater than this. Impairments on tasks that do not involve familiarity judgements attest to the fact that perirhinal cortex is involved in the greater role of knowing about objects, including, but not limited to, their relative familiarity.

This commentary attempts to reconcile the predictions of Aggleton & Brown's theoretical framework with previous findings obtained from experimental tests of laboratory animals with selective hippocampal lesions. Adopting a convergent operations approach, the predictions of the model are also related to human neuroimaging data and to other complementary research perspectives (cognitive, computational, psychopharmacological).

We applaud Aggleton & Brown's affirmation of limbic diencephalic-hippocampal interaction as a key memory substrate. However, we do not agree with a thesis of diencephalic-hippocampal strict dedication to episodic memory. Instead, this circuitry supports the production of context-specific patterns of activation that subserve retrieval for a broad class of memory phenomena, including goal-directed instrumental behavior of animals and episodic memory of humans.

Aggleton & Brown (A&B) propose that the hippocampal-anterior thalamic and perirhinal-medial dorsal thalamic systems play independent roles in episodic memory, with the hippocampus supporting recollection-based memory and the perirhinal cortex, recognition memory. In this commentary we discuss whether there is experimental support for the A&B model from studies of long-term memory in semantic dementia.

Experiments on memory-retrieval dynamics support the hypothesis that different mechanisms mediate processes of familiarity and recall (recollection). The minimal retrieval time for accurate recognition and frequency judgments is about 100 msec earlier than the minimal time to judge details such as modality or location. The difference is consistent with neurophysiological evidence. Retrieval dynamics data are also relevant to other aspects of Aggleton & Brown's target article.

Some of the assumptions underlying the arguments in Aggleton & Brown's target article are reviewed; discrepancies/predictions are pointed out in relation to human lesion studies. A&B's proposal is interesting, but it may require harder, confirmatory evidence before it can be considered to be all-encompassing.

An alternative to Aggleton & Brown's interpretation is presented suggesting that the perirhinal cortex and hippocampus mediate different attribute information, but use the same processes, supporting the idea of parallel processing based on attribute (visual object and spatial location) rather than process characteristics (item recognition and familiarity).

Aggleton & Brown predict that recognition and episodic recall depend on different brain systems and can thus be dissociated from one another. However, intact recall with impaired recognition has not yet been demonstrated if the same type of information is used in both tests. Current evidence suggests that processes underlying familiarity-based recognition are redundant with processes underlying episodic memory.

Fornix problem: Why do lesions of the fornix, which connects the hippocampus (HF) to the medial mammillary nucleus (MMN), often cause no deficits in tasks severely affected by lesions of HF or MMN? Solution: The direct HF feedback to antero ventral (AV) thalamus (MMN ⇒ AV [hArr ] HF), which is blocked by MMN lesions but not fornix lesions, is sufficient for nonscene-relevant consolidation.

Aggleton & Brown have built a convincing case that hippocampus-related circuits may be involved in thalamic amnesia. It remains to be established, however, that their model represents a distinct neurological system, that the distinction between recall and familiarity captures the roles of these pathways in episodic memory, or that there are no other systems that contribute to the signs of amnesia associated with thalamic disease.

The idea of distinct brain systems for the processing of episodic and other forms of memory is welcome. The two brain systems actually proposed however, appear to be stripped of further existing connections and could be integrated with one another. If integrating them, it seems more logical to propose one enlarged system of limbic structures whose individual components make partly different contributions to the forms of memory under discussion.

A hippocampal patient is described who shows preserved item recognition and simple recognition-based recollection but impaired recall and associative recognition. These data and other evidence suggest that contrary to Aggleton & Brown's target article, Papez circuit damage impairs only complex item-item-context recollection. A patient with perirhinal cortex damage and a delayed global memory deficit, apparently inconsistent with A&B's framework, is also described.

This commentary reviews recent evidence that some hippo- campal functions do not depend on perirhinal inputs and discusses how the multiple-process model of recognition may shed interpretive light on previous reports of DNMS reacquisition deficits in pretrained subjects with hippocampal damage. Suggestions are made for determining whether nonhuman subjects solve object-recognition tasks using recollective memory or familiarity judgments.