VARIABILITY OF RATE OF CULTURAL CHANGE 353
and with kite-shaped shields, the hypothesis of invariability in the tempo
of change would lead us, here too, to fly in the face of the chronological
facts by guessing that the Time-interval between Otto I (regnabat et
imperabat A.D. 936-73) and William the Conqueror (ducebat et regnabat
A.D. 1035-87) must have been notably longer than the Time-interval
between Majorian (imperabat A.D. 457-61) and Otto I.

In the three cases just cited, estimates based on the assumption that
the tempo of change is invariable fall wide of the true chronological
marks because in reality the tempo was not constant. In all these three
cases a spell of relatively slow cultural change was succeeded rather
abruptly by a spurt of relatively fast change. We may complete our ex-
position of the argument against the hypothesis of invariability in the
tempo of change by citing an inverse case in which a spurt of fast change
was succeeded by a spell of slow change.

Anyone who takes a synoptic view of the standard Western male (non-
military and non-clerical) dress as worn in A.D, 1700 and in A.B. 1950
respectively will see at a glance that the coat, waistcoat, trousers, and
umbrella of A.D. 1950 are merely variations on the coat, waistcoat,
breeches, and sword of A.D. 1700. By contrast, the doublet and trunk-
hose of A.D. 1600 are as different from the Western civilian costume of
A.D. 1700 as the Western military equipment of A.D. 966 is from the
Western military equipment of A.D. 1066. If a child ignorant of the dates
were asked to guess, from a series of pictures of celebrated Western
poets, which two of our three sets of costumes were separated from one
another by a Time-interval of a century and a quarter, and which two by
an interval of two centuries, the innocent child would assuredly guess
that the century and a quarter was the interval between Pope (natus
AJ>. 1688) and T. S. Eliot (natus A.D. 1888), and that the two centuries
was the interval between Shakespeare (natus A.D. 1564) and Pope (natus
A.D. 1688).

These cautionary tales are warnings against the danger of confiding in
an hypothesis of invariability in the tempo of change as a basis for trying
to estimate, not only the Time-interval between an Egyptiac Middle
Empire and New Empire, but also the length of time that it took for
successive strata of the debris of human occupation to accumulate on
some site whose history has to be reconstructed solely from the material
evidence disinterred by the archaeologist's spade, in default of chrono-
logical data furnished by independent, decipherable, and authoritative
written records.1 If, for example, we were tempted to compute the dura-

1 Such an attempt to -work out an absolute chronology by inference from the thickness
of strata of deposits has, of course, to be distinguished from an attempt to estimate the
relative age and duration of the strata deposited on different sites within the same broad
cultural field by comparing the likeness and differences between their respective con-
tents, on the lines of C. F. A. Schaeffer^ monumental and masterly Stratigraptee Cora-
parte et Chrvnologie de FAsie Occidentals (ro* et & Millhuares): Syne, Palestine, Asie
Mineure, Chypre, Perse et Caitcase (London 1948, Oxford University Press). In com-
paring the successive strata on different sites, Schaeffer is comparing entities that are
legitimately comparable. The error in method lies in assuming that the tempo of change
is constant and that it is therefore feasible to argue from the thickness of strata to dura-
tion of Time, or vice versa. Absolute chronological values cannot be assigned to strata
of debris with any likelihood of accuracy unless two conditions are satisfied. In the first
place the stratum that is to be datedmust contain some object bearing evidence of having

B Sfilfijct N