Further considerations regarding inhibitory processes, working memory, and cognitive aging.

The present study explored the relationship between inhibitory processing (as indexed by identity negative priming in a letter-naming task), working memory, discourse processing, and cognitive aging effects. Consistent with several other recent reports, this study found evidence of intact inhibitory processing, as measured by negative priming, in older adults as well as younger adults. This intact negative priming occurred in conjunction with diminished working memory span and impaired memory in the discourse processing task in the same sample of older adults, further arguing against the likelihood that declines in these areas are caused by impairments in the inhibitory processes measured by negative priming. The implications of these results for theories of cognitive aging and possible reasons for inconsistent findings regarding negative priming effects among the elderly are discussed.

Considerable research over the past two decades has been devoted to understanding the sources of age-related declines in certain cognitive capabilities. A number of these studies have suggested that a generalized slowing of processing speed may be an especially salient factor in cognitive aging effects in general (see Salthouse, 1996, for a recent review), but others have noted that decreased processing speed cannot account for all age-related cognitive changes (Nettelbeck & Rabbitt, 1992; Nettelbeck, Rabbitt, Wilson, & Batt, 1996; Salthouse, 1996). Therefore, it remains important to continue to evaluate other potential mechanisms.

One additional possible source of cognitive aging effects is a decline in the efficiency of inhibitory processes that may be critical in selecting information for further processing. The potential importance of inhibitory processes in cognition in general, and in individual differences in cognition in particular, has been the subject of intense scrutiny in recent years (see volumes by Dagenbach & Garr, 1994; Dempster & Brainerd, 1995). Hasher and Zacks (1988) focused attention on the possibility that inhibitory processes might be related to cognitive aging when they suggested that differences in the efficiency of inhibiting extraneous information from access to working memory might explain age-related declines in working memory. In Hasher and Zacks's account, inhibition prevented the access of goal-irrelevant information to working memory and suppressed or removed from working memory extraneous information that either had leaked in or had become irrelevant with changing goals. These inhibitory processes were thought to decline with age. Consistent with this suggestion, older adults have shown lower inhibitory functioning than young adults in studies using a variety of methods that are purported to measure inhibition. Age differences consistent with decreased efficiency of inhibitory processing have been found in stop signal (Kramer, Humphrey, Larish, Logan, & Strayer, 1994), Stroop (Houx, Jolles, & Vreeling, 1993; Salthouse & Meinz, 1995), text comprehension/memory for inferences (Hasher & Zacks, 1988), directed ignoring (Zacks & Hasher, 1994), and negative priming (Hasher, Stoltzfus, Zacks, & Rypma, 1991; McDowd & Oseas-Kreger, 1991; Stoltzfus, Hasher, Zacks, Ulivi, & Goldstein, 1993) tasks.

Of these measures, the negative priming paradigm popularized by Tipper (Neill, 1977; Tipper & Cranston, 1985; Allport, Tipper, & Chmiel, 1985) has been used particularly often to evaluate Hasher and Zacks's (1988) theory. Typically, this paradigm involves the presentation of a target stimulus and a distractor stimulus on pairs of trials, with the target identified as such by virtue of its color or location. Subjects are required to make a response to the target such as naming it, giving its physical location, or making a semantic classification. When the target stimulus on the second (probe) trial is the same as, or related to, the ignored distractor from the preceding (prime) trial, slowed responding typically is found, giving rise to the term negative priming Initial interpretations of the negative priming effect attributed the slowed responding to active suppression of the ignored item's representation or of the access from that item to action (see Tipper & Milliken, 1996; and May, Kane, & Hasher, 1995), thus making it seem a good candidate for assessing changes in the efficacy of inhibitory processes with age.

Following this idea, Hasher et al. (1991) contrasted negative priming effects in older adults and college students. College students showed significant negative priming effects, but older adults did not, consistent with the idea of a decline in inhibitory processing with age. However, further studies using negative priming measures have produced equivocal results: Some studies find clear age differences (Hasher et al., 1991; McDowd & Oseas-Kreger, 1991; Stoltzfus et al., 1993; Tipper, 1991), but others either find no age differences (Kramer et al., 1994; Schooler, Neumann, Caplan, & Roberts, 1997; Sullivan & Faust, 1993) or find that the presence or absence of age differences varies with procedural changes (Connelly & Hasher, 1993; Kane, May, Hasher, Rahhal, & Stolzfus, 1997; McDowd & Filion, 1995).

Evaluating the implications of this inconsistent pattern of results for Hasher & Zacks's theory is further complicated by the possibility that negative priming may not always assess inhibitory processes or may not adequately assess them. Neill and colleagues have proposed an episodic retrieval account of negative priming in which the slowed response to the target on the probe trial is caused by automatic retrieval of the prior processing episode in which a nonresponse for that particular stimulus was encoded rather than an inhibitory process (Neill & Valdes, 1992; Neill, Valdes, Terry, & Gorfein 1992; Neill, Valdes, & Terry, 1995). Park and Kanwisher (1994) suggested that negative priming occurs when the probe target differs from the item that was located in that position on the preceding prime trial, regardless of whether the item in that position on the prime trial was a target or a distractor. The merits of these different accounts, along with the original suggestion that negative priming reflects inhibit ory processes, continue to be debated (see Tipper & Milliken, 1996, and Neill & Valdes, 1996, for recent summaries).

Another development in the examination of inhibition and aging has been the suggestion that inhibitory processes are not a single resource that declines with age. Instead, discrete inhibitory mechanisms in a number of cognitive domains might be relevant to performance. However, because the mechanisms are discrete, each might be affected differently by aging. Thus, the relationship between inhibition and aging might vary as a function of the particular domain. Some support for this possibility has been obtained (Connelly & Hasher, 1993; Kane et al., 1997; Kramer et al., 1994).