THE present investigation deals with the behaviour of the time interval which occurs between the calling out of the stimulus-word and the utterance of the reaction-word. I term this interval simply "reaction-time," with the knowledge that it is made up of a composite series which can be deductively and ...
THE present investigation deals with the behaviour of the time interval which occurs between the calling out of the stimulus-word and the utterance of the reaction-word. I term this interval simply "reaction-time," with the knowledge that it is made up of a composite series which can be deductively and empirically reduced into numerous components. I forbear venturing upon an analysis of that kind, for the results would be just like hypotheses which had to be supported by anatomical data to which they were not entitled. The components of our " reaction-time" are only known to us in part, and in an exhaustive critique would prove extremely complicated, as Claparède's summary shows :
1. Claparède, L'Association des Idées, p. 275. The scheme is based upon Ziehen, Die Ideenassozialion des Kindes.
(1) Transference of the sound to the ear of the percipient.
(2) Nerve conduction to the acoustic centre.
(3) Recognition of the word (primary identification).
(4) Understanding of the word (secondary identification).
(5) Evoking of the induced presentation, i.e. the pure association.
(6) Naming of the presentation evoked.
(7) Excitation of the speech-motor apparatus or of the motor centre of the hand, if measured by means of the Morse communicator.
(8) Nerve conduction to the muscle.
A superficial consideration of these eight factors shows that we have only emphasized some of the most important factors. The innumerable possibilities of intracerebral processes have been in no wise exhausted.
So far as we know these components, they are of very short duration, the longest not exceeding 50σ. (Ziehen). Under normal conditions some of these components, e.g. the time for nerve conduction, excitation of the centre, etc., have a fairly constant duration. In any case, variations here take place within relatively narrow boundaries. Variations in identification time are greater, and greatest of all are the true association time and the time for the verbal formation of the reaction. In the association experiment the last factors have therefore greatest importance. Whoever has undertaken association experiments knows how 'wide are the limits of the reaction-times. In our experience times up to six seconds are not at all exceptional even among normal persons. This great difference of the times gives us a needful hint as to the methods of time measurement. So long as we have not sufficient knowledge of the causes of the variations, small differences in the times cannot tell us anything. We do not, therefore, require any complicated experimental conditions to measure times of one-thousandth of a second; we may quietly ignore slight differences so long as the causes of the greater differences do not escape us. Apart from the fact that complicated methods of measuring the finer intervals disclose nothing more than measurements with the one-fifth second stop-watch, there are weighty considerations against the use of mouth-whistles, trumpet-calls, or dark chambers. If Mayer and Orth1 regarded it as necessary to shut the eyes during the experiment to avoid all disturbing sense impressions, the apparatus mentioned certainly does nothing to simplify the experiment or prevent disturbing influences. In any case unpractised persons cannot be employed in these experiments if gross disturbances of attention are not to be risked. Finally, in mental patients the more exact time measurements are obviously excluded.
1. Zeitsch. f. Psychol., Bd. XXVI.
Measurement by the one-fifth second watch seems, therefore, not only quite sufficient but has stood the test in the investigations of several workers. Mayer and Orth  have worked with it, as have Thumb and Marbe, Wreschner, Sommer, and others. Caparède's view is that it suffices for all experiments in successive association. Besides simplicity in handling, the stop-watch has the especial advantage of disturbing the experiment least, and this fact was important in the experiments with the uneducated, who are prone to emotion.
2. Loc. cit.
3. Experimmtelle Untcrsuchungcn über die Orundlagen der spraehlichen Analogicbildung
4. Allgemeine Zeitschrift für Psychiatric, Bd. LVII.
5. Loc. cit., p. 261.
With the great differences in time that have to be considered, it is of little consequence that the times measured are all somewhat too large. Any one who has worked with this stop-watch knows that the apparatus is only approximately correct in its working, for the stop arrangement does not always put the hand exactly at the place at which the stop is liberated. To this must be added certain variations in the personal equation which may influence the measurement.