The Place of Statistical Methods in Modern Historiography*

• Author: Ernest Rubin
• DOI: http://doi.org/10.1111/j.1536-7150.1943.tb01079.x
• Date: 01 January 1943
• Published date: 01 January 1943

The Place of Statistical Methods in

Modern Historiography*

By ERNEST RUBIN

I

^E OBJECTIVE PURPOSE of this paper is to indicate certain

considerations towards a scientific history, by employing in

case illustrations some of the concepts of science and

as embodied in statistical methods. This paper

Only suggestive in character, no attempt being made to

5 a scientific or statistical theory of history. In sta-

tical methodology, however, there are certain elements of

y^ which should be brought to the attention of the his-

. "^^^n, particularly the historian of mass phenomena and

itutional developments in modern times/

sfore proceeding to the body of this paper, a word is

Pessary concerning the patterns of thought familiar to

^ematics and science, and the differences among the

ienis of mathematics, science, and history.

J ^^^ntific problems are distinguishable from those of his-

d'ff ^^ several unique ways. Perhaps the most important

|"ence is that of experimentation in science, which is im-

^ ie in historical analysis bounded by various ideological

and t^^^ ^^ ^*^ ^^^ limits and purposes.^ In science, theories

t»v V^^^^^^ses may be checked, revised, or destroyed in part

ne experimental method. Furthermore an important

V

fo/. y indebted to Professor Harold Hotelling of Columbia Uni-

.

'^

"As J '^'^'^''^'^'^ and suggestions related to the statistical aspects of this paper,

^l " *"''^''4 science no organism is pretended to be understood as long as its super-

ibd, so in history no movement by a mass of people can be correctly

^Conotn i ™3's 'S resolved into its component parts." Charles A. Beard,

D .

>

(new"^ri-•^"^'^^'^*^'°" °^ '^* Constitution of the United States," New York, lilac-

**6ly

sr>

I '^^'^")i I9i6, p. 2J3. Statistical methods are powerful tools in analyzing

'^' ^XVll

M*^

^' ^'^^"y'> "Indeterminism in Physics and History," Tie Social Studies,

• "^o, 8 (Dec. 19i£), pp. JH-J, for an interesting discussion of this point,

'Hoi J 193

194 The American Journal of Economics and Sociology

implication of the experimental method is control, which

would

seexn

to be lacking in even the most rigorous historical

analysis- And while the fruits of scientific inferences ofteii

contain predictability (though sometimes couched in the

language of probability), historical analysis in its present state

is chary of this element—if not hostile entirely to such an

idea.

The problem of mathematics differs from that of history

and science since the truths it seeks to discover lie in ^

matrix of postulation and logical analysis exclusively;

is,

the "facts" of mathematics are in part the axioms of

and several special assumptions and definitions, from

various complete algebraic or geometric systems may be o^'

duced. In a certain sense mathematics may be considered

a

an extension of logic.

While the work oi the pure mathematician is

strictly in a non-empirical realm, there are elements of

in the pattern of mathematical thinking for the

Mathematicians in higher analysis, when examining the

retical aspects of the calculus, for example, attempt .

strengthen the statement of the theorems. This process o

strengthening examines the mathematical and logical aspe^

of the theory, and by irriposing such criteria as "necessary

^

sufficient" conditions can delimit the periphery of the P^°^.

sitions. The net result is to strengthen the conclusion of

^

theorem, since it has been submitted to rigorous

proof,

*^^

the same time to understand what the theorem implies uo

given operations. , ^

It at once becomes obvious to the historian that the tn^^ j

matical mode and the scientific approach would lead to

results if applied in that form, by the historian to his

o-^

^

results if applied in that form, by the historian to hi ^

But the goal of the historian is not alien to that of the m^ .

matician and the scientist, since all seek the truth '^ *" j^

respective fields, and all are interested, when absolute tr^J