Welcome to Gaia! :: View User's Journal | Gaia Journals

So I'm going to repost it here.

Learning "How" and Learning "That" May Not Represent Separate Memory Systems

In the study of memory, a distinction has been made by many researchers and over many years between “knowing that”, which has been called explicit memory, and “knowing how”, or implicit memory (Cohen and Squire, 1980). The distinction was made after a series of observations of neurological patients that demonstrated that people with certain types of brain damage were impaired in their ability to recall facts or events, but whose ability to learn skills was left relatively intact (Corkin, 1984; Scoville & Milner, 1957). It is clear that knowing that and knowing how are different, but how do they differ? One possibility is that different parts of the brain perform different functions in the storage and retrieval of different kinds of memory.

In the late 1970s and early 1980s, several research teams investigated the memory phenomena observed in amnesic patients with the aim of finding a single basic memory process, above the molecular level, that would give rise to the observed pattern of impairment and spared function. However, most of these attempts faltered (Willingham & Goedert, 2001). The idea of independent memory systems was drawn from their failures and from the initial observations of amnesic patients’ capabilities and deficits.

One of the most famous observations that different parts of the brain are involved in different kinds of memory comes from Scoville and Milner’s 1957 study of surgical damage to or removal of parts of the hippocampus. The results of that study led the researchers “to attribute a special importance to the anterior hippocampus and hippocampal gyrus in the retention of new experience” (Scoville & Milner, 1957), especially given their patient H.M.’s nearly total inability to retain memories for events after the surgery. However, subsequent studies showed that H.M. was able to acquire and retain certain kinds of memory, most notably procedural memory, such as the skill to complete puzzles (Cohen & Corkin, 1981), identify fragmentary drawings and perform novel-familiar discrimination tasks (Nissen, Cohen & Corkin, 1981). This suggests that while the hippocampus (the structure removed in H.M.’s surgery) may be necessary for forming new memories for events and facts, encoding other kinds of memory does not require the hippocampus to be intact.

The evidence of this has come from numerous other sources. For example, in Cohen and Squire’s 1980 study, amnesic patients (whose memory problems came from a stab wound to the head, Korsakoff’s syndrome and electroconvulsive therapy) learned how to read words that were presented to them backward, through a mirror, as presented in Figure 1. Their performance was comparable to that of the controls. However, compared to the controls, the patients had very poor recollection of having seen any of the words they read in the course of the study. This suggests that the neurological foundations of being able to recognize words as familiar have been disrupted in these patients, but the neurological foundations of being able to learn to read English words when turned backward have been left more or less intact.


The roots of the idea that there are multiple memory systems lie fairly far back. As noted by Rovee-Collier, Hayne and Colombo (2001, p. 20-21), Renee Descartes explained his idea of how human beings are conscious by suggesting that there are two entities involved in human existence, the body and the soul, and that only the soul possesses the attribute of consciousness. (While this is not directly relevant to the question of how memory works, it does address consciousness, and conscious experience is defined in terms of memory.) Endel Tulving noted that Descartes also suggested the brain was involved in consciousness. Tulving went on to point out that while scientists have rejected dualism to conclude that consciousness is ultimately the product of brain activity, there is nonetheless a certain kind of dualism in modern memory research. He suggested that there are at least two types of memory, with the following question and observation:

Tulving then proposed that amnesia is a disorder of consciousness. He pointed out that in cases of amnesia, the subject generally suffers deficits in particular kinds of awareness of the world, but not in others; for example, a typical patient might not be aware of significant events from his own life, yet he is certainly aware enough of the world to hold conversations, as H.M. can (Corkin, 1984, or Newhouse, 2007). Within this frame of reference, Tulving suggested that “different kinds of consciousness, and different kinds of conscious awareness [including memories of different kinds], are assumed to represent inherent functional properties of those brain structures or mechanisms that subserve different kinds of memory” (Tulving, 1987). In short, Tulving made the statement that different kinds of memory are represented by different activities in different parts of the brain.

However, Tulving’s suggestion that there are fundamentally different types of memory reaches too far without a proper foundation in empirical evidence. Part of the reason for this is that the root of memory is the molecular changes in the structure and function of neurons, especially at synapses, based on phylogenetically old mechanisms such as the cyclic AMP pathway, a process that strengthens synapses. On a cellular level, all memory is probably based on these phenomena. Eric Kandel wrote, of learning in sea hares and fruit flies,

Kandel proposes that all memory ultimately results from the same set of phenomena – experiences triggering a series of chemical reactions in certain synapses that selectively alter the strength of those synapses.

The main problem with Tulving’s dual-system concept of memory is that having functions localized in different parts of the brains does not make the different memory phenomena into evidence for intrinsically different systems. It seems more parsimonious to assume that there is only one unified memory system, that needs certain circuits in the hippocampus to encode declarative and episodic memory, and which uses other circuits in other parts of the brain (the cerebellum, for example) for other tasks, such as learning mirror-drawing. Certainly it is now understood that the way different parts of the brain are used to recall a given memory can change over time; Bayley, Hopkins and Squire (2003), for example, found evidence that as memories become more remote with the passage of time, the brain circuits used to recall those memories are increasingly in the neocortex rather than the hippocampus or medial temporal lobe. However, this merely tells us about the phenomenon of memory as localized in different parts of the brain, and doesn’t shed any light on whether memory consists of one system or two. Given the lack of unambiguous data, unnecessary multiplication of entities should be avoided on general principles.