Life Sciences Vol . 5, pp . 17-26, 1966 . Printed in Great Britain.
Pergamon Preae Ltd .
LATEACY OF THE EYELID UCR DuRI1QG CC3iDITIONIIPG Irene 1lartin and A.H . Lsvey Institute of Psychiatry, Jniveraity of London, London .
3.x .5 .
(Received 8 October 1965) Interest in the fate of the UCR during conditioning has recently been renewed .
Studies have drawn attention to ohangea in UCR amplitude during
acquisition of the QR (1, 2, level of conditioning (4, 5) .
and these have been found in turn to relate to In our own work with eyelid conditioning we have
observed marked changes in UCR latency.
This report will present results of a
systematio study designed to investigate these ahangea and to sasses their generality .
The study is based an conditioning records selected from a large
scale project involving two levels of C3 intensity and CS-UCS interval .
provided an opportunity to evaluate these parameters, as well ae ease of~oonditioning, in relation to UCR latency ahangea. Method Apparatus
3 was seated comfortably in an adjustable dental chair
mounted in a sound deadened experimental room, and separated from E end tha apparatus by a heavily curtained panel .
Responses were recorded photoeleatri-
tally by a method which avoids arty form of mechanical interfersnoe with the eyelid, and which has been deaaribed in detail . eleewhere (6) . 250B ink-recorder, running at s paper speed of
mm,/seo provided a continuous
frictionlesa tracing of eyelid activity calibrated for each S to a deflection of 3 cn fnr full normal closure of the eyelid .
A separate channel recorded
onset and termination of C3 and JC3 . The CS was a tone delivered through conventional earphones by an audiogenerator.
A solenoid valve controlled the onset an8 duration of the JCS, a
LATENCY OF THE EYELID UC8
Yol . 5, No . 1
puff of dry oasipreseed air released through a rigid terminal jet, 70 mm in length and 1 ~ in diameter, directed at the vertez corneas from a distance of approzinatsly 1 .5 on .
The delivery delay of the UC3, measured by manna of a
modified condenser miorophone l, was 20 meec . and its mazimum intensity was reached 12 neeo . after impact .
All UCH latencies were scored from the point
of ~C9 impact. The signal for the masking tnek, described below, was given by a circular red light 1 .5 am in diameter, mounted in the centre of s 3~6" z 4~6" panel which faced the 3 at a distance of 4~, the chair being adjusted so that this light was at eye level . Procedure The conditioning procedure was imbedded in a simple vigilance teak requiring S to attend to the red light which came on 5 eeo. prior to the C3 onset, and terminated 5 sec. after.
He was told to ignore the distracting
tons and air puff, and to respond to the termination of the light by preening n hand switch . The ues of this masking technique virtually eliminates socelled voluntary response to the C3 . The C3 was a tone of 1000 cps at 65 db lasting 400 or 800 cosec. and overlapping a 60 cosec. air puff at nominal pressures of either 3 or 6 psi (which correspond to 150 or 300 mm~hg) .
Immediately prior to the conditioning pro
aedure the threshold for raflez blinking was determined for each subject against a scale of tinimal puff'preseurea .
The measure was the average
preeeure,aver three aeaending series, at ~iah a raflez blink of less than 200 nseo . latsnay said more than 1 ~ in magnitude was first elicited . following schedule was then admlnietered :
3 presentations of C3 alone, 3 of
UC3 alone, 3 CS alone and 40 paired presentations using intertrial intervals of 15, 20 or 25 seconds, arranged in an unsystematic sequence about a mode of 1.
Vfe are indebted to Dér. P . Bayer for constructing this equipment .
voi . 5, xo . i
OF R~E EYELID IICE
~y deflection of 1 mm or more from the stable baseline oocarring
between 25p mast . and the termination of the air gaff was snored ae a C8 . Sub ecte The Sa were psychologically naive normal adult male volnateere, age range 20 to 40 yeara2 .
They were oarefully randa~mised to the oonditions of the
original ezperiment representing two factors of puff intensity and inter stimulus interval .
In order to senate orthogonal representation of the parar
meter of conditionnbility, all reoorde were first categorised into three grouper those who had failed to oondition (lees than ioned poorly (less than adequately or well .
CRs), those who had oondit-
CEs), and the remainder who had oonditioned
3e were then chosen îrom each of these oategoriee in the
order of their acoession to the project, until the four oelle representing C3 intensity and CS-JCS interval had been filled with two records from each group to provide two replications of e 2 x 2
faotorial design .
The mean number
of CRa for the conditioning groups thus selected (1t ~ 8 in each) were 1 .62, 7.62 and 23 .88.
This procedure afforded an unbiassed comparison among the
three factors of puff intensity, interstimulua interval sad oonditioning level for a group of 24 Sa . Since ACS intensity and C3-JC3 interval would normally be ezpected to affect the level of conditioning, it is important to note that this design made it possible to study the relationship between UCR latency and condition ability independent of uCS intensity and CS-uCS interval . Results Res~onae measures :
The onset latency of the UCR was scored to the nearest
10 meet . for the three presentations (uCR alone) and for the 40 acquisition trials .
Für those trials on which a CR overlapred the JCR the point of de
pasture of the slope of the latter was scored as its onset .
In a few inatanoea
The authors would like to thank the Officers and volunteer Ss of the G.P .O . who assisted in this study.
LATENCY OF THE EYELID UCR
5, No . 1
in which the CR completely overlapped the JCR the latency cold not be
determined and these trials were disaerded. in
Figure 1 shows the wear. latencies
trial blocks of the acquisition aeries ? demonstrating a progressive
decrease in mean latency for each of the oonditioning groups .
The slope of
Figure 1 Decline of UCR latency over acquisition trials for three levels of oonditioning .
UCR latency upon ordinal trial number was chosen as the measu^e of rate of change, and was calculated for each S using standard regre^sior. techniques . An appreoiable and oonaistent tendency for the JCR to decrease over successive acquisition trials was demonstrated . slopes, ranging from -0 .0947 to
one positive, were not significant .
Of the 24 Ss, 22 gave negative
expressed sa maec .~trial.
The mean for ell Ss was -0 .593 "
proportion of variance accounted for by the linear trends ranged from 2.0',d to
with a mediar. value of
19 .6 .
The possibility was considered that these slopes might include some artefact due to the occurrence of CRs .
If, for example, CR occurrence Mere a~soc-
isted with shorter UCR latencies it would result in an apparent decT~ea~e i .
Yol . 5, No . 1
LATENCY OF THE EYELID UCa
latency over trials ea the probability of CRs increased.
To test this possi-
bility, mean absolute UOR latencies were assessed separate7,y, following the appearance of the first CR, for trials with sad withous CRs .
means were compared for individual 3e and for ,the group, and were found not to differ eignifiaantly . As e further check, individual elopes were recalculated .for the first subject in each cell (one replication) ezaluding all trials on which a CR occurred . The effect of omitting such trials was slightly to increase the negative direction of the elope and to decrease the variability of the aeseurea .
the effect, if any, of presence of the CR was to attenuate rather than to p~roduce the latency change under investigation. A ai.milar decrease in latency of the raflez blink to the air puff alone was also observed .
The mean IICR latencies over the 3 teat trials were 108.3,
93 .5 and 87 .0 m.seo . respectively, this difference being significant at the .005 level ( F 1 r 17 , 16 .59)'
this drop was uncorrelated either with
the elopes over trials in acquisition (r ~ 0 .103)~or with the drop ia~latenoy across the first three of these trials (r - 0.028) .
The mean latencies of the
latter were 98 .3, 92 " 5 éad 84 .6 m.aec . respectively . The return of the mean UCR latency to a level not signifiasntly different from that of the first, but differing from the last teat trial (Fl p
17 "52, r 2l ~
.005) is e puzzling feature of these results, since the acquisition trial
followed immediately on the teat trials, with only the three unpaired C3 presentations intervening .
The lack of a correlation between latency decrease
over teat and acquisition trials would seem to suggest that different ieohanieme underlie the decrease in UCR latency to paired and unpaired presentations of the JC3. Conditionability r
9nalyaie of variance of the slopes,, representing
decrease in SCR latency during 40 acquisition trials, showed they to be unrelated either to puff intensity or to CS-UC3 inteYwal .
For the conditioning
LATFI~C7 OF ~E ETELID UC$
Yol. 5, No . 1
groupr, howver, mean slopes of -0 .235, -0 .582 and -0 .963 differed si.gnifiaantly (l2~19 - 4"38 . p x.05) oorreeponding to the degree of conditioning as shown in lig~~e 1.
Thsrs were no significant interactions .
A correlation of 0 .529
(p x .01) was obtained between these elopes and number of CRs.
Ae e check on
the initial latency of the three oonditioning groups, average latency of the first three acquisition trials was calculated ;
thin was found to be unrelated
to degree of oonditioning, but significantly related to puff intensity (Fl 7 "24
p <.025) . The overall linearity of the trends led to the hypothesis that the amount
of decrease in UCR latency would effectively predict subsequent conditioning . This was tested in an analysis of verianos of the elopes upon trials preoeding the first CR .
Bean elopes for the conditioning groups from poor to good were
-0 .440, -0 .g90 and -1 .600, representing e slightly wider separation of the groups, though with no increase in significance
~ 3 .85, p x .05) "
oorrslation with umber of subsequent CRe was 0 .554 (P " x .01), indiaeting a substantial predictive element in the behaviour of the UCR prior to the appearanca of the first CR . Dissuasion This study gives formal support to an observation lade in our laboratory that the latency of the UCR deoreaeee during eyelid conditioning .
was to investigate the generality of~ this change in terms of three factors which might be ezpected to relate to the timing fluiationa of the UCR, viz. C3 intensity, C3-JC3 interval, and level of conditioning .
It therefore was a
necessary feature of . the design that the level of conditioning be kept orthog~ anal to the intensity and interval parameters .
The results show that changes
in UCR latency are not only related directly to level of acquisition but also predict conditionability . Heoauae oY the novelty of these findings a further replication was carried out on an entirely new group of 22 Se with minor procedural differences (e .g .
LATENCY OB THE EIELID IICB
CS-UCS interval 50p m.seo . eoquieition trials m~mbsring 30) . corded above were significantly~oonfiraed in all aspects;
The results re-
negative slopes of
JCR latency over trials ranged in this study from -1 .076 to -0 .0193 . The facts that the conditioning groups did not differ in initial absolute latency to either unpaired or paired presentations of the UC3 indioatee that initial sensitivity to the UC3 is not a asjor factor in the findings .
supported by the failure of the threshold aeaaure to diaoriainats conditioning groups . However, the olsar finding of a decrease in latsnay over the three test trials suggests that some faoilitatory process is occuzTing with IIC9 repetition . This decrease is not significantly related to oonditionab311ty, although the trend ie in the predioted direction. Rhen acquisition begins, the JCß appears restored to its original first test trial latency ;
one possible interpretation of this finding 1e that the
introduction of the novel C3 serves temporarily to abolish whatever faoilita tort' process we~ operative .
On subsequent eoquisition trials there is again a
rapid decrease in UCH latency which is especially Barked in good conditioners, althoti:gh this decrease ie not related to the observed decrease over teat trials . In considering mechanisms to account for these findings of decreased UC$ latency, several possibilities arise.
One ie that a senaitimation process
occurs in those neural structures which mediate the unconditioned blink, with the result that faster latencies are possible .
The finding of a decrease over
test trials can perhaps be ezplained in this way. Some other quite independent factor may be responsible for the UCH latency decrease which occurs during acquisition trials .
This could be learning of a
time discrimination, which is involved in CR acquisition, ceivably affect JCR latency.
(7i and could con
If an increasingly quick response to the JC3
implies learning by the S of a more accurate estimate of CS duration, then it is possible that a factor of timing efficiency may be reflected in the
LATENCY Ok THE EYELID UC&
behaviour of the UCR as well as that of the ~R .
Yol . 5, No . 1
Some indirect support for this
notion is seen in the fact that the UCR latencies of the poor conditioners were eonaiderably more variable that those of the good conditioners . This was reflected in the difference in aignificanee levels of the elopes for these groups, only one of the poor conditioners reaching the .005 level while only one of the superior group failed to da so . It should also be recalled that while elopes of UCR latency were uninflusnced by UC3 int~naity, the mean latenoy of uCRa during early acquisition trials were related to puff pressure in that the stronger puff intensity pro dunes shorter UCRe .
There is ample evidenoe that puff intensity affects level
of conditioning, which Rearan (8) suggests is mediated by the strength of the 0Cß rather than a hypothetionl drive variable . with this suggestion .
Our findings are in agreement
Irrespeotive of puff intensity and corditionability,
the first CR tended to appear at the point where JCR latenoy had dropped to within a range of 65-80 m.sec .
Ehrther, the fact that mean VCR latencies of
the three trials immediately preceding the appearance of the CR did not differ oong oonditioning groups or between levels of puff intensity, carries the interesting implication that some absolute range of UCR latency may provide the setting required for the emergence of the CR . In general, the behaviour of the UCR, in contrast to that of the CR, has received very little attention .
The finding of Kimble and his associates
(1961, 1963) that the amplitude of the UCR decreases as a flu~ction of the presenoe of the CS, and that this is related in turn to ease of conditioning bears some parallel to the present results.
The puzzling finding that the l3CR
latenoy of the first acquisition trials returned to the level of the first test trials, also confirmed in the replication study, would seem to impute some role to the CS .
FSirther, the fact that the phenomenon reported here cannot well be
interpreted as an inhibitory effect, and indeed carries implications of some exoitatory process, underlines Kimble~s contention that changes in the UCR
Pol . 5, Ao . 1
LATII~C7 OF ~ EYELID IICE
cannot be attributed merely to habituation. The eztent to which these and other changes is the behaviour of the IIC8 during conditioning ere independent of, or intwract with the developing C8 remains uncertain.
In our own work with eyelid oonditionlng w are inclined
increasingly to the view that the growth of the CS oocars concoa
whether this virr is jnsti-
Pied or not, there would appear to bs considerable justifiaatioa for a sore extensive inveetige.tion of the behaviour of the IICß during the oonditionlng process . ~bstraot Changes in IICß latency during eyelid conditioning wre inveetipted ed 24 3e in an orthogonal design involving two levels of CS intensity and ®-II08 interval and three levels of conditioning . Significant dsorsases is UCE latency over the aoQUisition trials wre found.
These were related to level of conditioning snob that the good con-
ditioners showed a eignifioaatly sore rapid decrease .
àhrther, the asount of
decrease prior to the first CB significantly predicted the level of ocuditioning. These findings were entirely supported in a further replication. Joknowlgd~sent~ The financial support of the Yedioal Heuaroh Connoil is gratellilly acknowledged . we are indebted to Professor H.J . ~senok for snppert and eaoouragement. Beferenoes 1.
SIIIBLE, G .~ . and OBT, J.ILP .
J. ~v . Pavohol. 61, 150 (1961v .
KI!®LE, G .A . and PEIQ~YPAC~, H.S .
PROBASY, wILLIeI[ F. Classical oonditioni 96 Appleton-Century-Crofts
J . Genet. Psyohol. 1~,, (1963) . .
(Bd. w .l . Proiraay),
LATENCY OF TFiE EYELID üCR
Yol . 5, No . 1
H'IIIB~L, H.D . and PEANYPACSER, H.3 .
J. Exp . Peychol.
6~e , 20 (1962) .
NICHOLL3, Y.F . and R~LE, G :A .
BLUFFISLD, R. and HOLLAI®, H .C . erimenta in liotivation . (Ed. H .J . Eyeenck.) Pergamon Preae, Oxford, 19 3 .
PROBA9Y, ".F ., SBEL, H.C . and THOMPS~P, D.D . (1963) "
RAZRAA, GREGORY. Psychol. Bull . ~, 1-(1957) "
J. Exp.~Psychol. ~,q00 (1964) .
J . Exp. PeSrchol . _66, 139