1. The non"conscious modulation of the processing, "retrieval or production of a mental item by prior exposure to specific information on that item or on items associated with it.

2. The presentation of a "stimulus or the induction of a change that prepares the "system for functioning.

'Priming' in the current memory literature almost always refers to a specific protocol or type of memory (definition 1) rather than to the more general concept (definition 2). This contemporary use was introduced by Cofer (1960), on the basis of earlier observations

(Storms 1958; see also Wohlgemuth 1913; Williams 1953). An experiment by Storms (1958) illustrates what was meant. Storms asked a group of students to study a list of words, A, which elicit high-frequency associations, such as 'eagle' (which typically elicits 'bird') or 'hill' ('mountain').1 He then presented the students with another list of words, B, which included the words that are high-frequency responses to the words in A,but themselves do not ordinarily elicit words in A. For example, although 'eagle' elicits the response 'bird', the converse is not usually the case. Storms found that the production of words in A as responses to words in B was significantly higher when list A was encountered just previously. Cofer (1960) termed this use-dependent augmentation of associative strength as 'priming of associations'.

Over the years the use of the term 'priming' has been extended to include additional protocols and meanings. The current major "taxonomies of priming are based on two types of interrelated distinctions: 'direct' vs. 'indirect', and 'perceptual' vs. 'conceptual' (Richardson-Klavehn and Bjork 1988; Roediger and McDermott 1993). In direct priming, also termed 'repetition priming', the item presented in the training ('study') "phase is identical to, or is composed of fragments of, the item to be produced in the test phase. In indirect priming, also termed 'associative priming', the item presented in the study phase is associated with the item to be produced in the test but not identical with it. Direct priming is mostly 'perceptual'; it is modality specific and relatively independent of the meaning (semantics) of the item. It may last for many weeks. Indirect priming, which persists for much shorter periods, is mostly 'conceptual' or 'semantic'; it is dependent on the meaning of the item, and relatively independent of the sensory modality.

Repetition priming is where the frequency, speed, or accuracy of response is facilitated as a consequence of prior exposure to a particular stimulus (Tulving and Schacter 1990; Ochsner et al. 1994; Wiggs and Martin 1998). The most commonly used repetition priming "assays include 'perceptual identification', 'word completion', and 'lexical decision' tasks (Schacter 1987; Roediger and McDermott 1993). On a typical perceptual identification task (Jacoby and Dallas 1981), the subject is exposed on the study phase to series of words. Each word is flashed on the screen for a fraction of a second only. In the test phase, series or lists of words, including both the previously presented and 'new' words, are presented and the subject is requested to identify them. Priming is reflected in the increase in the accuracy or in the speed of identification of the previously presented words compared to the 'new' words. On the word completion task (Roediger et al.

1992), the subject is presented in the test with either word stems (e.g. AIR... ), or word fragments (e.g. _ R_ _ I _ G), and instructed to complete them with the first word that comes to mind; priming is manifested as the preference for words that have been presented on the study phase (e.g.'AIRCRAFT','PRIMING'). On the lexical decision task (Meyer and Schvaneveldt 1971), the subject is required to state whether or not a particular letter string is a legal word, for example, ARDUBOK, BELABOR, GARGOZOM (the middle one is a word, the two others are pseudowords, unless psychologists or molecular biologists have invented something since this has been written). Priming in this case is reflected in the decreased latency to make the decision on the second presentation of the letter string relative to the first. The examples above refer to visual presentation of verbal material; similar tasks could be performed with other sensory modalities and nonverbal material.

Commonly used associative priming assays include word associations, category production, and general knowledge priming tasks. On the word association task (Storms 1958; Shimamura and Squire 1984), a word presented in the study session results in preference for an associated word, e.g. stem-flower; the study by Storm (1958), mentioned above, is but an example. On the category production task (Srinivas and Roediger 1990; Gabrieli et al. 1995), the subject is presented with items from a certain category (e.g. animals), and later asked to produce as many as possible category exemplars (e.g. 'lion', 'elephant'); priming is manifested by a "bias toward production of the previously studied category exemplars. On the general knowledge task (Blaxton 1989; Vaidya et al. 1996), presentation of a word will enhance the production of an answer to trivia questions related to that word; for example, presentation of the word 'Jerusalem' in the study phase, will facilitate the production of the correct answer to the question 'What is the capital of Israel'.

Priming is a hot topic in research on human memory. There are four main reasons for this popularity. First, priming using verbal material is useful in the analysis of the perception, processing, and production of language. Second, priming, similarly to "transfer, provides a window to the processes and mechanisms of retrieval and to their dependence on the conditions and processes of encoding in "acquisition. Third, priming, a non "declarative memory, illuminates distinctions among memory systems ("taxonomy; Squire et al.

1993). And fourth, priming protocols are used to tap residual memory in "amnestic patients. It was noted almost a century ago that past information in 'global' amnesic could sometimes be recovered if the patient is provided with clues to that information (see in Williams 1953). It is now evident that perceptual priming, and possibly some capabilities of conceptual priming, are preserved in amnesics (Warrington and Weiskrantz 1968; Backer Cave and Squire 1992; Schacter and Buckner 1998; for conceptual priming that is impaired in amnesics, see Vaidya et al. 1996). As 'global' amnesia is known to occur as a consequence of damage to structures in the medial temporal lobe and the diencephalon, it follows that perceptual priming is not critically dependent upon the integrity of these areas. The introduction of noninvasive "functional neuroimaging has expended tremendously the ability to identify brain regions that subserve priming (Schacter and Buckner 1998;Henson et al. 2000;Yasuno et al. 2000). In a nutshell, perceptual priming is correlated with reduced neuronal activation in modality specific neo"cortex, especially in higher-level processing areas. Conceptual priming is correlated with a similar change in multiple neocortical areas, including the prefrontal cortex. At the cellular level, repetition priming is thought to be correlated with a decrease in neuronal response in cortex with repeated stimulus presentation ('repetition suppression', Desimone 1996).2 However, the possibility that in some circuits repetition priming involves experience-dependent synaptic facilitation, e.g. "long-term potentiation, should not be neglected (Milner 1997; for a candidate mechanism, see Frey and Morris 1997).

It is important to re-emphasize that 'priming' covers heterogeneous processes, which are expected to be subserved by multiple circuits and mechanisms in the brain.3 The common denominator to all these processes is the nonconscious use-dependent facilitation of the processing and retrieval of an item.4 "Sensitization also facilitates future processing, retrieval, and "performance, but is nonspecific, whereas priming depends on the specificity of the item(s) presented in the study phase. In some of its properties, such as the independence of "conscious awareness, priming resembles "habit and "skill. In others, it resembles declarative memory: it provides the brain with specific, discriminative, and precise information about events in the world. Tulving (1983) speculated that in performing priming experiments, we tap into memory capabilities that had emerged in phylogenesis after procedural memory systems but before declarative memory, and have played an important part in the life of early hominids ("Homo). Were those ancestors of ours knowledgeable about the world but not consciously aware of it the way

Prospective memory we are? Were they a bit like global amnesics? Even if the role of priming early in the evolution of our species remains a mystery, analysis of the distinction between brain systems that subserve priming and those that subserve declarative knowledge might provide us with clues to the identity of brain circuits that subserve consciousness.

Selected associations: Acquisition, Retrieval, Skill, Taxonomy, Transfer

1For another 'classic example of the use of word associations in unveiling intriguing properties of memory, see *false memory.

2This finding is also discussed in the context of 'habituation.

3This raises the question whether priming is a memory system, or a property that cuts across different memory systems. The *zeitgeist is to classify it as a memory system (*taxonomy), but it is unlikely that repetition and conceptual priming are subserved by the same system. This classification problem is not unique to priming; take, for example, 'classical conditioning. It is considered a system of procedural memory, but trace conditioning is declarative, whereas delay conditioning is not. The take-home message is that taxonomies should be taken seriously only as much as they promote new concepts and research. 4Those cases in which the prior exposure results in inhibition of processing, are termed 'negative priming'.

0 0

Post a comment