You are on page 1of 5

llrif1iIJ Jour/wi ~f P~j'thf)/~,fI.

J' (1993), 84,249- 253 Printed ill Gmll Brit"i,j © 1993 The British Psychol )gical Sudety


Clock* this!

An everyday example of a schema-driven


error m memory

Christopher C. French+

DcpaTl7tlClll oj P[,)'choil1gy, Gof.olJmithf College, UniTlcrrity of London, NelP Cross) London J' 14 6N W, K

Anne Richards

Depatttmnr of PJ_Jehologr, Birkbe.k College, Londo»

n everyday example of a distortion of me11l(;}J'Y is reported. objects were presented with a standard clock with numbers represented by Roman numerals, Of rhe subjects asked to d.raw the clock Ftom memory, the majority mistakenly represented the four as 'IV' rather rhan as 'IllI '. The latter represents the standard form of Roman numeral representation upon docks, but nor elsewhere, it made no difference to performance whether Or not subjects were warned ill advance thar they would have to draw the dock from memory, Subjects asked simply to copy the dock in full view dld nor make the C(rOI. The findings are interpreted ill terms of schema cheery.

The last 20 years of memory research have differed from most of the previous 80 in that the study of everyday memory has been recognized as being of great importance (Cohen, 1989), One of the most striking findings from this body of recent research is that memory for common everyday objects is often remarkably poor. For example)

ickerson & Adams (1982) showed that American subjects performed extremely poorly 00 tasks involving the recall and recognition of a U one-cent coin. Morris (1988) reports that British students performed even worse in retognjzing the correct appearance of a 10p. piece. Previously, Morton (1967) had shown that not one of SO British subjects could correctly recall the positions of the letters and numbers on the British telephone in the days when both appeared upon the dial

Schema theory is often inv ked to explain many of the phenomena of everyday memory_ In Cohen's (1989) words 'It can account for the fact that many of our experiences are forgotten, or are reconstructed in a way that is incomplete, inaccurate, generalised, or distorted. Schema theory emphasises the role of prior knowledge and past experience, claiming that what we remember is influenced by what we already know, According to thi theory, the knowledge we have stored in memory is organised as a set of schernas, or knowledge structures, which represent

'" 'clock I ... ,J British s,"'ng meallJ11g [0 look at I ... j' (from P: 105 of B,..",.r', Dirti."ary rif 201h.Cimtll'Y Pbrase and Fubk., London: Cassell, 1991,. edited by D. Pickering; fl.. l,,~c~ & E. Martin).

t Requests for reprints,


Christopher C. French and .Anne Richards

the general knowledge ab ut objects, situations, events, or actions that has been acquired from past experience' (p. 71). Schema theory provides the theoretical basis for the study reported below.

The inspiration for the study was supplied by an incident involving the :first author and the second author's daughter, Lucy Richards, a couple of yeats ago, when the child was aged about eight. In. the course of visiting her grandparents, Lucy's attention was caught by the Roman numerals upon a clock face in the ro lTI. The conversation then proceeded something like this:

Lucy: On the clock, why does 'V' come after 'lill'?

CCF (without looking at the clock): It doesn't say 'illI'. It says 'IV' for fou.r. Lucy: It doesn't. Look.

CCF (looking at clock): Incredible! You'd think clock-makers of all people would know Roman numerals! But this is how it should be. (Shows his wrist-watch) Would you believe it, they've got it wrong on here as well!

In fact, as we now realize, virtually all docks with Roman numerals feature 'lllI' instead of 'IV', a fact which will probably come as a surprise to most readers, According to Clutron, Baillie & Ilbert (1982), thi is due simply to aesthetic considerations as ' ... tradition has favoured the strictly incorrect lIII, which more nearly balances the equally heavy VnI' (p. 95). However, it is obvious that schema theory provides a powerful explanation of this failure of everyday memory on the part of the first author. Even though all adult must have been exposed to thousands of examples of clocks and watches featuring Roman numerals, the fact hat four is represented as 'IIIl' is not relevant to telling the time and therefore geL1erally goes unnoticed. Instead we tend to misremember the appearance of such timepieces in line with our schematic knowledge of the Roman numeral system.

The study reportedinvestigated this phenomenon under controlled conditions, by asking subjects to draw such a dock. Subjects either copied the clock directly or else drew it from memory. Of those subjects drawing from memory, about half were informed in advance that they would be required to do so, and about half were not so informed. Under the viewing conditions employed, .it was predicted that most, if not aIL, f the subject copying the clock directly would correctly represent the four as 'IIII'. It was predicted, upon the basis of schema theory, that those subjects copying from memory would show a tendency to represent the four as 'IV') in line with their schematic knowledge of the Roman numeral system. It was of interest to examine whether or not the forewarning of the memory task would serve to reduce this tendency.



The subjects were 41 volunteers (31 female), attending an open day for candidates applying for B c courses in the Department ofPsycbology, Goldsmiths' College. Ages ranged from 17 to 46 years, with 29 subjects aged 18 or less.


Subjects were run in groups of six Or seven in order to ensure that ail subjects bad a good clear view of the stimulus dock. The dock used in this study was a. smnchrd dock made by Samuel Bishop of

Clock tbis!


London, with clear Roman numerals on the face. The fOL1:r was, e,f course, represented as . IIIT '. The clock was sec at 1 Q minutes past seven.

Subject groups were allocated to ODe of three conditions as [0110';.\,5:

(/) Condition A. The experimenter rold these subjects, • I am gomg to show you a clock which I want you to examine visually for one minute'. The cluck was rhen removed from. view and subjects were issued with pencils, erasers and paper (this took exactly two minutes). They were then asked to draw the clock as accurately as po sible. from memory and informed rhan they would be allowed six minutes to do so (after five minutes, they were warned that only one minute remained). This condition will henceforth be referred to 3S the "frprise "MIlIOry tt1sk. Fourteen subjects were run .. in this condition.

(il) Condition It The experimenten told these subjects, 'I am going to show you a dock which J want you to examine visually for one minute. Then I will ask you to draw it from memory. The clock itself will be removed. You will be allowed sis minutes'. Following the examination period, the clock wa removed from view and subjects were issued with pencils, erasers and papet(aga.iu, taking two minutes exactly). Subjects were allowed six minutes to draw the dock (will" It warning after five minutes). This condition will henceforth be referred to as forewarnet! ultlnory tosk. Fourteen subjects were run in this condition also.

(iit) Conditioll C. In this condition, the subjects were issued with pencils, paper and erasers ill advance, and the dock remained in fuji view for the six-minute copying period. The experimenter told the e subjects, 'I would like you ro draw a picture of this clock. You have six minutes'. objects were again warned when five minutes had elapsed. This condition will henceforth be referred ro as the copy condition. Thirreen subjects were run in this condition, although one subject was excluded because her incomplete drawing after ix minutes did not include any representation of the four on the dock face,

AU subjects were debriefed Cit flIOSS~ following completion of data collection.

Results and discussion

Each drawing was inspected to see if the four on the clock face had been represented correctly as'IllI' or incorrectly 'as 'IV', Results for the 40 subjects providing usable data are p1:'e ented in Table 1.

Table 1. Subjects representing the four as 'IV' and <UU' under the three conditions of the experiment: surprise memory task (SMT), forewarned memory task (FMT) and copy task (CT)




'IV' -nn-

9 5

10 4

o 12

As can be dearly seen, no subject in the copy task misrepresented the lIII' as 'IV'. Under the clear viewing conditions used in this study, the influence of subjects' schematic of the Roman numeral system wa not strong enough to affect their actual perception of the clock in the copy condition (although it is clearly possible that suchan influence .might be demonstrated under different viewing conditions). It is apparent, however, that subjects' drawings from memory were strongly influenced by their schematic knowledge. In the surprise memory task condition, nine of 14 subjects misrepresented the 'lIIl' as 'IV'. Of even greater interest, however, is the fact that the subjects in the forewarned memory task


Chri.ftophcf· C. French and .Anne Richards

condition performed no better, with 10 of J 4 subjects making the same mistake. i\ chi square test revealed that the conditions differed significantly (X2 (2) = 15.65, P < .001). and examination of cell frequencies shows that this reflects the much better performance of the subjects in the copy condition compared to those in the memory conditions,

It is perhaps surprising that those subjects in the forewarned memory task condition did not perform any better than those in the surprise memory task condition, although both groups clearly performed much worse than those subjects in the copy condition. Two possible explanations suggest themselves. First, it is possible that at least some subjects in the surprise memory task condition anticipated that their memory for the appearance of the clock might be tested, thus reducing the difference between the two memory conditions. After all, the subjects were aware that they were taking patt in a psychology experiment. It is possible, therefore, that if this possibility were reduced, say by presenting images of the dock incidentally during performance of another task, the number of memory distortions in a surprise memory task might be even higher than the impressively high number noted in the current study.

A second possible explanation attempts to account for the rea on that the forewarned memory ta k group performed so poorly in comparison to the copy group, despite the knowledge that they too would have to draw the clock. The subjects in the copy condition would clearly adopt a cognitively economical strategy of simply glaacil1g at the clock whenever they wished to examine any particular feature. Hence, all subjects copied the' Hfl ' correctly. u bjecrs in the forewarned memoty task condition, however, knew that they would have to tty to commit to memory many details of the clock, including the correct time, the maker's name and the place of manufacture (both written 0:0 the face), the shape of the hands, the orientation of the numbers, and so on, in addition co the actual Roman numerals. But why, then, did most of them not notice the apparently 'odd' four (' odd " that is, in terms of their schematic knowledge)? If they had noticed, they would surely have remembered this unexpected feature. It seem likely that most subjects probably simply examined a couple of numbers in order to take note of the general style f the numerals and relied upon their schematic knowledge to fill in the details during the actual drawing.

Of course, some subjects in (be memory conditions, around a third, actually represented the' nIT' correctly .. It is likely that some of these subjects noticed for the Erst time in their lives. as it result of this study, that fours arc represented in this unusual wayan clocks with Roman numerals. Some were perhaps already aware of this oddity. However, it is worth pointing out that had the experiment involved simply asking subjects t draw a representation of a typical clock face with Roman numerals from memory, without first allowing them to study such a clock face for a whole minute ill he context of a psychological experiment, there would no doubt have been an even greater number of rni representations. The influence f schematic knowledge on memory is thus shown to be powerful indeed in this particular everyday example.

Clock this I



Clurton, c., Baillie, G. H. & Ilberr, C. A. (1982). Brillm',r Old Cloi'kJ (mdW'll't~he:s and .il1eir M(lker.r, 9dl

ed. London: Bloomsbury Books. -

Cohen, G. (1989). ivluflf1ry i11 tbe Real lf7orld. LondDIJ: Erlbaurn,

Morris, .P. E.. (19gB). Expertise mel everyday L:BCmOrj'. In 1\1.. M.. Gwncberg, p, E. .Mnrris & R. N, Sykes (Eds), Practical AspeC'tJ' oj J,,[I'I"~ry: CNTrel11 Retm1'~b and [niles, vol, 'L Chichester: Wiley.

Morron, J. (1967), A singular lack of mcidenrsl learning. Nature, 215, 203-2Q4.

Nlckeno!1, R. S .. & Adams, i\L J. (1982). hong-term memory for a common object. In U. Neisser (Ed.), MeUlo']! O&s811led: Refl'«lIIbmi~g if! Natnra] C011/j1XtJ.. San Francisco: Freeman.