The first component of the dual-store model, the sensory register , holds incoming information long enough for it to undergo very preliminary cognitive processing. 3 For example, if you’ve ever
played at night with a lighted sparkler—one of those metal sticks that, when lit, emits fiery sparks for a few minutes—then you’ve seen the tail of light that follows the sparkler as you wave it about. This trail doesn’t exist in the air itself; it’s the result of your sensory register saving the light for a short time after you’ve seen it. As another example, when you sit in a classroom for an hour or more, your attention will at some point almost inevitably wander away from the instruc- tor. You may have noticed that when you tune back in to what the instructor is saying, you can often recall two or three words that the instructor said right before you tuned back in. Virtually
everything the instructor has uttered has probably been recorded in your sensory register despite your mental absence from class, but, alas, only those last two or three words are still there when you decide to be mentally present again.
Characteristics of the Sensory Register
Let’s look specifically at three characteristics of the sensory register: capacity, forms of storage, and duration.
Capacity As far as psychologists can tell, the sensory register has a very large capacity. For instance, even 6-month-old infants seem to record a good deal of what they see, albeit only very temporarily (Blaser & Kaldy, 2010; Neisser, 1967).
Forms of storage Information appears to be stored in the sensory register in basically the same form in which it has been sensed: Visual input is stored in a visual form, auditory input in an auditory form, and so on (Coltheart, Lea, & Thompson, 1974; Cowan, 1995; Howard, 1983; Turvey & Kravetz, 1970). At this point, information hasn’t yet been understood or interpreted by the learner. In other words, the sensory register holds information before any significant encod- ing occurs. 4
Duration Information remains in the sensory register for only a very brief time, but measuring its exact duration is difficult. A key problem in studying the sensory register’s duration is that when we ask people to report or in some other way process something they’ve stored there, the information automatically moves on to working memory and so is no longer in the place where we want to study it!
3 You might also see the terms sensory buffer, iconic memory (for vision), and echoic memory (for hearing). 4 As we discovered in Chapter 2 , different parts of the brain seem to be responsible for processing different kinds of sensory input. We can reasonably guess, then, that the sensory register isn’t a single structure located in a particular spot in the brain. Rather, different areas may act as sensory registers for different sensory modalities.
In an early study, George Sperling (1960) designed an ingenious method to assess the duration of information in the sensory register. Recruiting adults to participate in his experiment, Sperling presented displays of three rows of four letters and digits each; following is a typical display:
7 1 V F
X L 5 3
B 4 W 7
Each display was presented for a fraction of a second, and then participants were asked to recall either one particular row of symbols or all 12 symbols. When asked to recall a single row, people were able to do so with 76% accuracy; because they weren’t told which row they would need to
recall until after the display had disappeared, they apparently remembered approximately 76% of the symbols they had seen. Yet when asked to recall all 12 symbols, they could do so with only 36% accuracy. Sperling’s explanation of these results was that most of the symbols were stored initially but faded from memory before the participants had a chance to report them all. In a follow-up experiment, Sperling varied the amount of time that elapsed between the display and the signal indicating which row was to be remembered. People could recall little of a display after a delay of more than a quarter of a second.
Judging from Sperling’s early results, as well as from the results of more recent research, it appears that the duration of a complete visual stimulus probably remains in the sensory register for less than a second, although parts of it may remain for up to 2 seconds—perhaps a bit longer if no new visual information comes in (e.g., Gold, Murray, Sekuler, Bennett, & Sekuler, 2005; G. R. Loftus & Loftus, 1976; Sligte, Sholte, & Lamme, 2009). Auditory information is more likely to last for at least two seconds and often lasts even longer, with louder stimuli lasting longer than quieter ones (C. L. Baldwin, 2007; Cowan, Nugent, Elliott, & Saults, 2000; Darwin, Turvey, & Crowder, 1972; Lu, Williamson, & Kaufman, 1992).
Why might auditory input last longer than visual input? One possible explanation is that a major source of auditory input—human speech—can be understood only within its sequential context (Dahan, 2010; Grossheinrich, Kademann, Bruder, Bartling, & von Suchodoletz, 2010; Wingfield & Byrnes, 1981). For example, consider this sentence:
I scream for ice cream.
You can interpret the first two words as either I scream or ice cream . Only when you hear the third
word— for —can you begin to interpret the first two words accurately. The task of understanding
speech, which is frequently filled with temporarily ambiguous sounds, is often easier if we can hold those sounds in memory in an uninterpreted form until we receive additional, clarifying information. As we human beings have acquired a greater capacity for language, then, evolution may gradually have given us a greater capacity to retain uninterpreted sequential auditory input. Two factors probably account for the rapid disappearance of information from the sensory register (Breitmeyer & Ganz, 1976; G. R. Loftus & Loftus, 1976; Lu & Sperling, 2003; Sligte et al., 2009). First, interference may be a factor: New information coming in effectively replaces—
and thereby erases—the information already there. Yet even without new input, existing infor- mation in the sensory register seems to quickly fade away, or decay , over time. In any case, in
most instances people don’t need to store information there for very long. Important information is often going to be processed sufficiently that it enters working memory. Unimportant informa- tion, like junk mail, is probably best thrown away.