.

'Nk 1-2
��,!f·
�
[ 41]

,,,�L, ��
:�
i�
::��
�
CHAPTER III
t��
I�1 WHAT ARE "HISTORY" AND
"HISTORICAL SOURCES"?

The Meaning of "History"

� THE ENGLISH word history is der_ived from the Greek
111
''!·
noun l<nopla, meaning learning. As ·used by the Greek
�
�f CD1TSCH,AU< 0 LDvtS, ( (qGq). U11der-stav-dlvi53 h 1s--io<JJ ; philosopher Aristotle, history meant a systematic ac­

r
� count of a set of natural phenomena, whether or not

t
�� Q pdme:I'"- d? hi<;tvr"icv.l me.¼cd :· New �o,i< ·:.
chronological ordering was a factor in the account;
A. A-, Kinopf, and that usage, though rare, still prevails in English
�
� in the phrase natural history. In the course of time,
'
�\

however, the equivalent Latin word scientia (English,

�
� science) came to be used more regularly to designate
�
I
non-chronological systematic accounts of natural phe­
I nomena; and the word history was reserved usually

'
� for accounts of phenomena ( especially human af­
� fairs) in chronological order.
By its most common definition, the word history
�
t now means "the past of mankind." Compare the Ger­
''I
man word for history - Geschichte, which is derived
from geschehen, meaning to happen. · Geschichte is

�
q�
--� that which has happened. This meaning of the word
I history is often encountered in such overworked
phrases as "all history teaches" or "the lessons of his-
J;. t ory. "
u requires only a moment's reflection to recognize
that in this sense history cannot be reconstructed.
The past of mankind for the most part is beyond re-
1 t� 1!1:1·
<J

42 UNDERSTANDING HISTORY

with the best mem­

,; "msTORY"
'·. AND "msTORICAi souRc-Es" 43
A vulgar pre judice exists against "subjective" knowl­
'·' ,,�
,,,
. . ;,f.:i� call. Even those who are blessed in the life edge as inferior to "objective" k nowledge, largely be­
"i their own past, si nce
:�,, ories canno t re-create
.
. =»� ust be events, person
s, words, cause the word "subjective" has also come to m ean
:�J of all men there m
;i�:1 made no impression "illusory" or "based upon personal considerations,"
tho ughts, places, and fancies tbat
:1!r1 at all at the time the
y occurred, or have since been and hence either "untrue" or "biased." Knowledge
t,
·]i�f forgotten. A fortiori, the experie
nce of a generation may be acquired, however, by an impartial and judi­
records or whose cially detached investigation of mental images, pro­
lon g dead, most of whom left no
been disturbed by cesses, concepts, and precepts that are one or more
r ecords, if they exist, have never f tot3.l steps removed from objective reality. Impartiality and
eyori d the po ssibility o
i! ---rth e historian's touch, is b
. ;�J t 7
./l,Q_CAx-51/2.LW recollection. T�e reco nstruction of the total
� "objectivity," to be sure, may be more difficult to ob­
:1u ,rl .,t;Lc,
(;(-r-

� i-/iJ sJh_us
. e_g.:iaLcl his to ria n tain from such data,. and hence conclusions based
Ell
J1!1· J m_EI__kil1d,-al.though....itis _fii well is una ttain­ upon them may be more debatable; but such data
:! t
---:-
, /A, - ,,,J::ii;comes a goal they k now full
If"'u.,---v'-<
utiilr
if/­
�h
,i-t;:itt and conclusions, if true, are not n ecessarily i nferior
able. to other kinds of knowledge per se. Jne word subiec­
m:"��f "Objectivity" and "SU-b jectivity" ii::
; ' tive is not used h ere to imply disparagement ill__illl)!

l!i Sometime:; obj ects like ruins, parchm en

surviv�_fr-mntbepast. Oth erwise, the
ts, and SQ.ins
£acf�(}j_istqry
sort, but it does imcly the necessity f or the applica-
tio� of special kinds of safeguards aga nst err
�:0 < f�
:j}
_ �
�,,,\".- re are facts gJ
a!��d frora testimoE_Y and therefo
i� Artifacts as Sources of History -lry� -
.

!�0t �g. They c�n, _fe lt , � d, heaE_d, or � . r ·1 J,X.aM=-'

c or r_�p..r.e­ 0 nly where relics of human happeni ngs can be
.�....., i,,,
�-
'fi.;�.i- �Pi sm_tlkcL They may be said to be �
r1o1 sentative of som etl1 in _ th3t once was
real, but they found - a potsherd, a coin, a ruin, 3.�manuscrip.t, a
tfa�\.� ,· J·1 other wor ds,

i'j :ft
ow . I book, a portrait, a stamp, a piece of wreckage, a stran d
:- have no o i ective reality of t heir n n
oria n's mind of hair, or other archeological or anthropologicaLr.ec
they exist or:ly i n the observer's or hist
Ji.

mains - d o we have objects other than words that

(and thus may be called "subjective "). � g
d
of acquirin th e historian can stu dy. These objects, however, .are
obj ecti V1:!Y (that is, with the intention
owledge i pen dent of never the happenings or the events themselves. If arti-
detached and t ruthful kn
nde

il
be_an o_b­ facts, they are the results of events; if written docu-
o ne's personal reactions), a thin g must first
IV' \i:
---· an independent exis tence_m1tside
jec t; it must have ments, they may be th e results or the records of events.
---· ----=-- =-- _____::.:...:.....- do not have \Vhether artifacts or documents, they are raw ma-
·, thenuman rni,id . Recollections, however,
most of his- terials ou t of which history may be written.
existence outside the human mind; and
t ory i s based up on recollect
ions - that is, written or -1'•!}
T o be sure, certain historical truths can be derived
spoken testi mony. immedi ately from such mate rials. The historian can
 44 UNDERSTANDING HJ STORY ":aisToii.i·" · AND "HISTORICAL souRcEs" 45
'f�J­
·i!f·
·1,r. discover that a piece of pottery was handwrought, that
ti"jti a building was made of mortared brick, that a manu­ Historical Knowledge Limited
by Incompleteness of the Records
',-{'"•f;

·.·:.,�\�1''l:<i1· script was "�n in a cursive hand, that a painting

. �'<;·
·t.l was done in oils, that sanitary plumbing was known Unfortunately, for most of the past we not only
..;:·'f1P
!
l
in an old city, and many other such data from direct have no further evidence of the human setting in
:,;/ ,.

;:: �w,j�i·
_ ..
,
!,
observc1tion of artifacts surviving from the past. But
such facts, important thou h the are, are not the es­
which to place surviving artifacts; we do not even
have the artifacts. Most human affairs happen with­
sence o t Le study of history. The historian deals with out leaving vestiges or records of any kind behind
. �ITT.f the dynamic or genetic (the becoming) as well as the them. The past, having happened, has perished for­
•i�t
b: {. . ,_, static ( the being or the become) and he aims at being ever with only occasional traces. To begin with, al­
1�Jvf

:)t fi
'•1 iI_!:!=erpretative (explaining why and how things hap­
pened and were interrelated) as well as descriptive
(telling what h2ppened, when and where, and who
;
though the absolute number of historical writings is
staggering, only a small part of what happened in the

-mt
�i
past was ever observed. A moment's reflection is suffi­
tfi
took part). Besides, such descriptive data as can be cient to establish that fact. How much, for example,
.mf: !; l [,,{,,V,,U-VA-­
derived directly and im·mediately ifrom surviving arti­ of what you do, say, or think is ever observed by any­

1·t-· : �1
./f.'- ffe � ,, fac�s are only a small part of the peribds to which they one (including yourself)? Multiply your unobserved
;,�;?/ belong. A historical context can be given to them only actions, thoughts, words, and physiological processes
if they caitbe placed in a huG1an setting. That human by 2,000,000,000, and you will get a rough estimate
beings lived m the brick building with sanitary plumb­ of the amount of unobserved happenings that go on

t
•�l.
ing, ate out of the handwrought pottery, and admired in the world all the time. And only a part of what
the oil painting that were mentioned above might was observed in the past was remembered b:t--those
�· !ti perhaps easily be inferred. But the inference may just who observed it; onlup _ �as remembered

;t;J
as easily be mistaken, for the builciing might have was recorded_; on�art__ of what was re_cordcl...has

!
hen a stable, the piece of pottery r.:iight have been
from a roof-tile, the painting might have been a
hidden-away relic with no admirers whatsoever; and
an infinity of other suppositions is possible. Without
further evidence the hum:m context of these artifacts
can .:1e·1et be :i:ecaptured with any degree of certainty.
-=-
Sl_!�d; only a part of _'0'hat h�yived has COJI!e
to the histmians' attention; oply a part of what has
come to their attention is credible; onlµpa.r.t-oLwhat
is c_ redible has been grasped; and only a parJ; of what
has been grasped can be expounded or n�te1..by
th� histornn. The whole history of the, past (what
has been called history-as-actuality) can be known to
him only through the surviving record of it ( history-as­
record), and most of history-as-record is only the sur-

.. ,
 ,-�-r:,I•
!l,i
!�''

Ii
'.f't t,
46 UNDERST A.i'lDING HISTORY
\j
viving part of the recorded part of the remembered
�� f
part of the observed part of that whole. Even when
the record of the past is derived directly from archeo­
logical or anthropological remains, they are yet only
the scholars' selected parts of the discovered parts of
the chance survivals from the total past.
In so far as the historian has an external object to
I' study it is not the perished history that actually hap­
pened (history-as-actuality) but the surviving records
of what happened (history-as-record). Bistory can be
told only from history-ascrecord; and history as told
,, (spoken-or-written-history) is only the historians' ex­
:;1
:i pressed part of the understood part of the credible
part of the discovered part of history-as-record. Before
the past is set forth by the historian, it is likely to have
gone through eight separate steps at each of which
some of it has been lost; and there is no guarantee that
i
what remains is the most important, the largest, the
,......
most valuable, the most represeutative, or the most
(i enduring part. In other words the "object" that the
historiau studies is not only incomplete; it is markedly
variable as records are lost or rediscovered.
History as the Subjective Process
of Re-creation
From this probably inadequate remainder the his,
!,
torian must do what he can to restore the total past of
mankind ..He has no way of doing it but in terms of
his own experience. That experience, however, has
:1
taught him ( 1) that yesterday was different from to­
! day in some ways as well as the same as today in
"HISTORY" AND "HISTORICAL SOURCES". 47
others, and ( 2) that his own experience is both like
-'Ii
' '' and unlike other men's. It is not alone his own mem­
ories interpreted in the light of his own experience
that he must try to apply to the understanding of
historical survivals; it is the memories of many other
people as well. But one's own memories are abstract
images, not realities, and one's reconstructions of
others' memories, even when reinforced by contem­
porary records and relics, are likely. to be even more
abstract. Thus the utmost the historian can grasp of
history-as-actuality, no matter how real it may have
seemed while it was happening, can be nothing more
'
i than a mental image or a series of mental images
based upon an application of his own experience, real
and vicarious, to part of a part of a part of a part of a
part of a part of a part of a part of a vanished whole.
In short, the historian's aim is verisimilitude with
regard to a perished past - a subjective process -
rather than experimental certainty with regard to an
objective reality. He tries to get as close an approxima­
tion to the truthabo.J1Lthe_pas.Las....const�..DLcorre.c:
hon of his mental images will allow, at the same time
recognizing that that truth has in fact eluded hirn
forever. Here is the essential difference between the
study of man's past and of man's physical environ­
ment. Physics, for example, has an extrinsic and whole
object to study - the physical universe - that does
not change because the physicist is studying it, no
matter how much his understanding 9f it may change;
history has only detached and scattered objects to
study ( documents and relics) that do not together
 48 UNDERSTANDING HISTORY "mSTORY" AND "HISTORICAL' SOURCES" 49
make up the total object that the historian is study­ method) 1 the historian endeavors to reconstruct as
mg - the past of mankind - and that object, having
it �uch of the past of mankind as he can. Even in this

largely disappeared, exists only in as far as his always
incomplete and frequently changing understanding
of it can re-create it. Some of the natnral scientists,
such as geologists and paleozoologists, in so far as the
objects they study are traces from a perished past,
f
·limited effort, however, the historian is handicapped.
He rarely can tell the story even of a part of the past
,f "as it actually occurred," although the great Ger­
fi man historian Leopold von Ranke enjoined him to
do so, because in addition to the probable incomplete­

I�-
f, greatly reseffible historians in this regard, but differ
�! " ness of the records, he is faced with the inadequacy of
,! from them, on the other hand, in so far as historians the human imagination and of human speech for such
have to deal with human testimony as well as physical an "actual" re-creation. But he can endeavor, to use

I
�.;
, ✓

traces. a geometrician's phrase, to approach the actual past

'!!
Once the historian understands his predicament, "as a limit." For the past conceived of as something
!/
his task is simplified, His responsibility shifts fro_E1 that Hactually occurred" places obvious limits upon

I·
i'
the obligation to acquire a complete knowledge of

rF;
the kinds of record and of imagination that he may
tne irrecoverabie past by means of the surviving evi­ use. He must be sure that his records really come
dence to rthat of re-creatmg a verisimilar image of as from the past and are in fact what thei_seem to be
) mucho: the past as the evidence makes recoverable. t ;�dthat his imagination is directed toward re-cre�n
The latter task 1s the easier one. For the historian arid not creat10n. These limits distinguish history from
r-­ t'
history becomes only that part of the human past fiction, poetry, drama, and fantasy.
:1 which can be meaningfully reconstructed from the
) ·, I
available records and from inferences regarding their Imagination in Historiography
setting. The historian is not permitted to imagine things
_,
Historical Method and Historiography �-- that could not reasonably have happened, For certain
,i
.;
Defined � purposes that we shall later examine he may imagine
g
Th.LRrocess of critically examining and analyzing ·_\
1 Some confusion arises from the use of the term historical
the records -md survivals of the oast is here called method by practitioners in other disciplines ( economics ·-and the-
pistorical method. The imaginative reconstruction of 6Togyespecially) to mean the application of historical data and illus­
i th�J@St from the data derived by that process is called
trations to their problems. It will simplify our discussion to restrict
:f the term to the method by which historical testimony is analyzed
'l
:_j
historiowaphy (the writing of history). By means of
_historical method and historiography (both of which
f<i authentic and credible data. Courses by historians in'.'fmtorical
method,'' however, generally include not only instruction in such
analysis but also the synthesizing of such data into reliable historical
.j h are�uently grouped together simEJy as historical
expositions and narratives.
'f. l·
'
J .. l
!
i 1·
11
Li
f I
 .! ;'

50 UNDERSTANDING -HISTORY "HISTORY'.' AND "HISTORICAL· SOURCES" 51

things that might have happened·. But he is frequently
History of Historical Method
required to imagine things that must have happened.
For the exercise of the imagination in history it is Historical method, however, not only can be made
impossible to lay down rults except very general ones. the subject of rules and regulations; for over two thou­
�r
It is a platitude that the historianYiliQ__k_n._ows. coc.­ sand )'ears it has been. Thucydides, who in the fifth
�pora!]'.'._ life best will und.ers...tan.d_pg_s._Uife_b_es.t. century B.c. wrote his famous history of the Pelopon­
·l:� ,.·:,:I Since the human mentality has not changed notice­ nesian War, conscientiously told his readers how he
ii A
!: Ii� ably in historic times, present generations can under­ gathered his materials and what tests he used to sepa­
:.1

•-�i·) ,., stand past generations in terms of their own experi­ rate truth from fiction. Even when he· invented
�;l �! ence. Those historians can make the best analogies speeches to put into the mouths of contemporaries,
[(."'.:
;J.nd co!J.trasts who have -·the greatest awareness of he tried to make them as like the originals as his

n1:1
�J ):1
po�sibje analogies and contrasts - thc1t is, the widest
range of experience, imagination, wisdom·, and knowl­
''1-Ui
:.•, !/!
.i i 1 ; edge. Unfo,.tunately, r;.o platitude tells how to acquire
a wide range: of those desirable qualities and knowl­
edge or how to t,ansfer them to an understanding of
the past. For they are not accumulated alone by pre­
""l
(_{ i;;
cept or example, industry and prayer, though all of
these rr,ay heip. Ancl so historiography,2 the synthesiz­
j -1.r
� l ; :; ing of historical data into narrative or expositions by
writing history books and articles or delivering history
\).
lectures, is not easily made the subject of rules and
regulations. Some room must be left for native talent
sources of knowledge permitted. He hoped to conform
both to the spirit of the speaker and the letter of the
speech; but since stenographic reports were not avail­
able, he had sometimes to supply the speaker's ·.vords,
"expressed as I thought he would be likely to express
them." 3
Since Thucydides' day, many historians have writ­
ten, briefly or at length, upon historical method. Out­
standi;1g examples are _!:-,ucian, Ibn Khaldun, Bodin,
Mably, Voltaire, and �though sometimes their
studies have dealt with the scope rather than the tech­
i
r niques of history. With Ernst Bemheim's Lehrbuch
f

and inspiration, and perhaps that is a good thing. But
since precepts and examples may help, an effort will
be made ( see especially Chapters VII-XII) to set
Ii
ti l' forth a few of them.
··jt�;i·
� 2 Confusion arises here too from the fact that historiography is
il1:! sometimes used to mean the critical examination of history books
already written, as, for example, in college courses on "histori­
t
It der historischen Methode und der Geschichtsphiloso­
phie ( 1st ed., Leipzig, 1889), the modern and more
academic discussion of the subject may be said to
have begun. Since Bernheim's exposition a number of
other textbooks have been published. Although none
of them surpass his masterpiece, peculiar merits in­
tended for particular kinds of readers are found in

,J[!i
J '.,
ography." 3 Thucydides Translated into English by Benjamin Jowett, I
. (Oxford, 1900), 16 (Bk. I, 22).
,:-/!:

\Jt
II Ii !

; ij ;:
!\ii,
 !.i .t)
:
i. ::''
52 UNDERSTANDING HISTORY
"His'i'o:ky" AND "msTORICAL souRcEs" ·-53
some. Notable examples are the Langlo:s and Seigno­
bos volume for Frenchmen; the Johnson and the or numismpticaLma.terials, he has to depend fargely on *j
museums. Where they are official records, he may have �
Nevins volumes for Americans; the Harsin and the
to search for them in a�chives, courthouses, govern- �
;) Kent booklets for younger students; and the Wolf,
mental libraries, etc. Where they are Erivate papers
the Hockett, and the Bloch and Renouvin books for
-r';
•ii
�JA...-.,._

not available in official collections, he may have to

.. ::'. students of specialized fields of history.
hunt among the papers of business houses, the muni-

I•'
r l :li In all of these works and literally dozens of others
rnent rooms of ancient castles, the prized possessions
' tf
.

j1j like them there is a striking ctegree of unanimity2.e-

of autograph collectors, the records of parish churches,
J Ui gE:iling the metho.ds of '1J.storical analY§.is. For our
etc. Having so�e subject in mind, with more or less
..

:J)i: purposes these methods will be considered under four

definite delimitation of the persons, areas, times, and
i:i ;r'. n : headings: ( 1) _t_!ie selection of a subject for investiga­
�
functions (i.e., the economic, political, intellectual,
;/ (';�, tion; (2) the collection of probable sources of ip­
(� ;.;l J,; I_ forr.nati9n on thaLs��t; ( 3) _ the �xamination ..?f
diplomatic, or other occupational aspects) involved,
he looks for materials that may have so�e bearing
. _-..� ::·I r: � th�_p..uJ!,:eS for genumen�ss (either �n whole_ or m
-. ;J }Li)��:"'_ p�E!.1._and (4) t1.:_e extracbon of credible parbcu�rs upon those persons in that area at that time function-
,·{25
:-,·· 'i:; �(!.,_��om the sources_, (or parts of �ourcesLproved_ �n�­ ing in that fashion. These materials are his sources.
The more precise his delimitations of persons, area,

l
' . ' �[�n,vu?J:"'.�, m�The synthesis of the particulars thus denved 1s
time, and function, the more relevant his sources are
L/(J),d/.C{./{IL�L"-'l.r,,,fi.istoriography, about which there is less _ unanimity
�- .

likely to be ( see Chapter X).

� among the textbooks For purposes _ of clan _ ty we shall
1} :� r:11,(,_ '".,tpLl
!;j.) I)�. � Tt.Vhave to treat analysis and synthesis as 1f they were
_.
The Distinction between Primary
'L( {vnri � discrete processes, but \�e sh�ll see that at various
, :! O stages they cannot be entirely separated. t and Other Original Sources
Written and oral sources are divided into two kinds:
Scurces
l primary and secondary. A primary source is the testi­
I! mony of an eyewitness, or of a witness b� any other o.f
The historian's problems in choosing a subject and
collecting information upon it (the latter sometimes
f
I, ___ ____________
th� senses,....,_or of a mechanical device..Jike....the_di.c.ta-

I,
I.

dignified by the Greek name of heuristics) will be qis­ ph011e - that is, of one who or that which was Eieseg_t
{)I l
-.'i. l� f"'�
cussed in Chapter IV. Histor1cal heuristics do not dif­
fer essentially from any__other bibliog@,P-hical exercise
at ·the events of which he or it tells (hereafter called
simply eyewitness). A secondary source is_ the testi.:.
:r 1Ii',:l: in so far as printedbooks are concerned. The histor��n, rriony of anyone :who is not an eyewitness-=--· that is,_Qf
one whowasnotpre� the._e..v:en.�
,f
"[ ;;;: however, h�to use many_ materials that are not in

/iil�ti�/\
rooks. Where these are a�<::_heolggica), epigraphi_£al, tells.A primary �ource_ 11_1_ust_tb_Y.� .Q.i!Y.� ..b.een �Qduced
bya·:.-�9!1t�mpora_ry of :�1?.�-���EJS it f!arrates. It does

rt I;,!
 : o,
'"
1
i ,·:
1�
. 54 UNDERST.AJ.'l"DINC HISTORY "HISTORY" AND "HISTORICAL SOURCES" 55
'i not, however, need to be original in the legal sense of have been lost ( in the sense that Livy is an "o�iginal
f the word original • - that is, the very document ( usu- source" for some of our knowledge of the kings of
• ,·.,i
,\
ally the first written draft) whose contents are the
subject of discussion - for quite often a later copy or
Hi"i Rome). In using the phrase historians are frequently p(l,VY'=�
guilty of looseness. An effort will be made to use it �
here only in the two senses just defined. � )">,,r( 6-::, cv,...,r��
I·..,, i-

'.I!;'ii! a printed edition will do just as well; and in the case ff)·"
ii
::.
i;. ,,,, of the Greek and Roman classics seldom are any but Primary sources need not be original in either of - 2:"v�"'::.
• -,('."

;r :,i:: later copies available. these two ways. They need be "original" only in the
.

:I /\)i , "Original" is a word of so many different meanings sense of underived or first-hand as to their testimonL
ko>-...,;(...._

�=�
::J:; ik ')7!.M-i-u,jj that it. would have been better to avoid it in .precise
'1 �ll; �lit/1,.L.a�- ' ·:)1l This point ought to be emphasiz�d in order to avoid

;,rm'.u �
gt�)� -. 0 : h 1sloncal d'iscourse. I t can be, and frequent1y 1s, used
to �enote fi:e different cotiditions of� document, all of
_
which are important to the histonan. A document
I,
-;
confusion between <?F�g_inal sources and Q_rimary
sources. The confusion arises from a particularly care-
lessuse of the word original. It is often used by his-
f.f :�!; ._. ;e
J!)�/1 may be called "original" ( 1) because it contains fresh l torians as a synonym for manuscript or archival. Yet a
;�1�
;J (,i,,J,,,�fand creative ideas ( 2) be�ause it is n?t tra�slated
.
moment's reflexion will suffice to indicate that a manu-
?fit q.,t::.v from the language m which 1t was first wntten, , 3) be­ script source is no more likely to be primary than a
'.

! \·:--:;-- . 11.,,,..,,....,,J2. cause it is in its earliest, unpolished stage, ( 4) because printed source, and that it may be a copy rather than
;J!, . 1

��1
'j '\_ CV its text �s the approved text, u�modifi�d ,md untam- the "original." Even where it is a primary source, it
. .
j'�) �u;[ pered with, a�d ( 5) b�ca�se 1t 1� the earliest available may deal with a subject upon which earlier informa-
(D.
>
�ource of the mformabon 1t provides. These five mean­ tion is already available. Hence a manuscript source is
}P! !1� �
. ...

:- '\:l �-Ul-d-'Yl. � ,.,.. mgs of the word may overlap, but they are not syn­
3,,
net necessarily "original" in either of the two relevant
>� :.:i i.�Q.rtfVl ...b .., onym..ous • senses of that word. It should be remembered that the
__
, ·.:,j
'· !,j'i'
Unfortunately, the phrase "original sources" has be- historian when analyzing sources is interested chiefly
J/!J'. t\y.J'1.L-gv�V come c?mmon among historia�s, and it is desirable �o in particulars and that he asks of each particular
i\) ·c�-rw��' define 1ts usage accurately. It 1s best use� by the h.is­ whether it is based on first-hand or second-hand testi-
',: .l:17\ 1..i \,aC,,rl.,,.,itonan _ m only two_ senses -- ( 1) to des�n�e a source, ·' mony. Hence it makes small difference to him whether
..'ff[: ���f/ unpolished, uncop 1ed, untranslated, as :t. issued from a document is "original" in the sense of "as written by
irH �J,,.,t/�i the the hands of the authors ( e.g., the ong mal draft of_ l�
··i"h ..t-
its actual author" or a copy, except in so far as such

nnl
J/hi� 't Magna Carta) or ( 2) a source that gives the earh: originality may aid him to determine its author and
_

JHifv
01,'tJ �st avail ble i,nfor1:1atio (i e., the origin) regardi11g
cl,._;rt., therefore whether it is primary or, if secondary, from
� � : .
e.J- f th_e qu_ es_ _t 1_ 0n _ _ un
_ der mvesbgabon because earlier sources · what more· independent testimony it is derived. Stu­
;_ �_, _ _ _
Jj:!.jl]l (),V<f',._.'P,$� _ 4 Cf. J�hn H. Wigmore, Student's Textbook of the Law of Evi­ dents of history readily depend upon specialists in edi­
;11 //J ; �,JlAIYl-al.xf" dence ( Chicago, 19 3 5), pp. 22 5-6. torial skills and archival techniques to publish collec-
:l i!·/- 1i!i
!ft
·-II""
�a
!�kr i-1

'f�I� nn
i!J!.
f � Af /
.., ti•/!
. :'\lP..
 71 -'.:

r
I'i �I

)1
56 UNDERSTANDING HISTORY �HISTORY" AND -"msTORICAL-SOURCES" - ·5
i' ,,
,;.
:i tions of manuscripts and are willing to use them m are trustworthy not because of the book or arficle or
l
' ,·:.
I,
printed form. '�.fr '. Cs•/
report they are in, but because of the reliability of the
narrator as a witness of those particulars. This point
�-·;•·-:::
t, jt 1;:f'.
Ii Primary Particulars Rather than Whole
!i
., Primary Sources Sought I �tt
·i -
..(7
_ 2·
will be elaborated later (see pp. 139-44).

: 1;1!
As has just been indicated, the historian is less con­ :: r
.;, � .{ ·The Document
j:
cerned with a source as a whole than with the particu­ � ;�.
i)i:i ItJi The word document (from docere, to teach) has
ij:hj lar data within that source. It is easy to conceive of a
also been used by historians in several senses. On the
'/ii! source essentially primary that will contain secondary ....

[�l
one hand, it is sometimes used- to mean a written
(and therefore less usable} data. The general who
source of historical informat�on as contrasted with
writes a communique thereby provides a source that
oral testimony or with artifacts, pictorial survivals, and
may be for the most part primary but for many details -tl
;I!- archeological remains. On the other, it is sometimes
secondary, because he must necessarily depend upon
.,,.,i1i� reserved for only official and state papers such as trea­
..�-'i/l: iJ;
' ! : !!
his subordinates for information regarding much that
ties, laws, grants, deeds, etc. Still another sense is con­
'1.• he reports. The newspaper correspondent may, like
tained in the word documentation, which. as used by
Aeneas, tell about things "all of which he saw and
I

./ �J'.i,'!! the historian among others, si@ifi.es any process qf

.s �:�·
part of which he was" and yet may aiso have to depend
i''I
��:. proof based upon any kind of source whether written,

i-0,!1 ··i ;
upon "an official spokesman" or "a source usually con­ :;, .. ;.�
I . oral, pictorial, or archeologi_gl. For the sake of clarity,
sidered reliable" for some of his information. The -; � .1'

:Iii-· it-seems best to employ the word document in the last,

/i.i.-·1;1 careful historian will not use all the statements of such }t�· ,

rW:!/ military communiques or newspaper dispatches with

equal confidence. On the other hand, should he find,
the most comprehensive meaning, which is etymo­
logically correct, using written document and official
document to designate the less comprehemive cate­
:f'i!' !i
';Ji! as he frequently does, that a book that is essentially
�f
gories. Thus document becomes synonymous with
secondary (like a biography or even a work of fiction)
r:1 •i source, whether written or not, official or not, primary
) contains, for example, personal letters or touches of
'-1':'/. or not.
directly observed local color, he may well use them as
·-1.,:· 11 ifi
L 11
first-hand evidence if they are genuine and relevant. ,11
,t,, i
A'f
'.1,.1 ,r·
Sources, in other words, whether primazy_or__s_e.c­ Th e "Human" and the "P ersonal" Document
y:::1
,{-.;,j ondary, are important to the historian becau�_they The human document has been defined as "an ac- �
di,ru,,.,,,,-,._,,_.,.
)1.-1 f>1j1}i co,E!ain primary partic�!ars (or at least suggest leads count of 1ndividual experience which reveals the indi­
to primary particulars). The particulars they furnish vidual's actions as a human agent and as a participant
:Jfiij
:1nl
l;r1 :f
11 ;lI,[,:, -�
1
/1!•: 111
,,!·'.•l•/1
�.r ..
:;,r:rl1
. Wti
.-.,,f-"[i•!
,!f,)il
I\.
!t
�,-�\..
 r-f!irarw
'" t!,
� �i !/ "HI�TORY" AND "HIST9_RICAL SOURCES" _59_
f ';/'. 58 UNDERSTANDING HISTORY
'ffr � I ;)I
'j;!: degree of subjectivity in these documents that distin­
in socic1l life." 5 The personal .document has been de­
i;
':j
'l'i fined as "any self-revealing record that intentionally or guishes them from ·other documents. The best exam­
:j
1
.1 1:
ples 8 seem to be documents written in the first person

;,/;\tJ!I
unintentionally yields information regarding the struc­
ture, dynamics and functioning of the author's mental - like autobiographies and letters - or documents

;i1,·
life." 0 The first definition is by a sociologist and em­ written in the third person but describing human re,

!t�·�'.
actions and attitudes - like newspaper accounts, court
1:·!i; phasizes "experience ... in social life" as an element
l .:�,

records, and records of social agencies.

.. \'ff;,1!11:ir of tr.e human document. The second definition is by a

-;1:t
psychologist and emphasizes "the author's mental life" ·! �
To the historian the difference between first-person
;�,, and third-person documents is not of major signifi
as an element of the personal document. Yet the words .
tit :·.

[i
human docum,znt and personal document have been cance. That is true for at least three reasons. ( 1) Often
-'' ·:i1· an apparently third-person document is in fact first.
used interchangeably. The two kinds of documents
1

seem to have one essential characteristic in common;
a human, persqnal reaction to the events with which
they deal. To both sociologist and psychologist it is the
,! .., 5 Herbert Blumer, An Appraisal of Thomas and Znaniecki's 'The
:�,· ·-r?
\;! i
person ( as, for example, the Memoires of Lafayette or
The Education of Henry Adams). ( 2) Genuinely
. third-person documents in so far as they are "histor­
icable" 9 must ultimately rest on first-hand observation
(whether by the author or by someone consulted by

<'!)!' Polish Peasant in Europe and America' ( "Critiques of Research in

·, ;1. the author). (3) Every document, no matter how

10
the Social Sciences;" Vol. I; New York, 19 39), p. 29.
6 Gordon vV. Allport, The Use of Personal Documents in Psy­
:·-i: thoroughly the author strove to be impartial and de­

ft
chological Science (New York, Soci.,l Scie:r,ce Research Council,
;'.);. 1941), p. xii.
7 Robert Redfield, "For-:word" to Blumer, p. viii. Cf. Allport,
pp. xii-xiv. Allport says that for the psychologist methods of evalua­
·,;_:; tion differ for first-person and third-person documents. They revolve
around "sources of material, observer reliability, and techniques of
presentation." For the historian, who as nearly as possible limits his
elementary data to primary particulars, these: are likely to be quanti­
tative rather than qualitative differences. That is td say, a participant
in a battle will probably have more nu,ncrous first-hand data to give
j:·{ than a newsp3pcr correspondent (who, incident�lly, may be less mis­
/_;i. taken than the particip�nt). Still, a first-person account by a par­
ticipant is valuable, as evidence, only for the particulars which that
/;d:
lr,ij participant gives on first-hand testimony or for leads to first-hand
testimony; and ;: third-person account of the same battle by a news­
,ir,.
;·•
tached, must exhibit to a greater or lesser extent the
•..:11{
if:··
author's philosophies and emphases, likes and dislikes,
and hence betrays the author's inner personality.u Ed­
ward Gibbon's Decline and Fall of the Roman Em­
pire, Johann Gustav Droysen's Geschichte Alexanders
des Grossen or Hippolyte Taine's French Revolution
8 Allport, p. xiii; Dlumer, p. 29.
9 I have had to invent this word to designate "capable of critical
examination by the historian." Please note that it is not a synonym
for true, reliable, or probable, but means only subject to inquiry as to
credibility.

l�i?
.:!/_::; 10 See note 7 above.
paper correspondent is valuable, as evidence, only for the same kind 11
c,f data. Allr- ort agrees that "the first-person and third-person docn­ Cf. Havelock Ellis, Dance of Life (Boston, 1923), pp. 8-12,
where the differe!lt interpretations of Napoleon by H. G. Wells and
;J,,.I.
mcr.ts ... both deal with the single case nnd on this question will
Elie Faure are attributed to the d.ifference lietween Wells and f<\un;.
1 stand or fal1 together." See also Allport, pp. 19-20.

q_:;11::;1r:L.
fil1!r }
if;
i;·,. iv::/
:,,,.,..... Ul! t,
...,,,. .. ''lll'1
.,.,, ft f":
il

�- flfl!
t:lt�ilti
 .,1;:1 60
-�/!ri
UNDERSTANDING !IISTORY "msTOR_y" AND "HISTORICAL SOURCES'..'. .61

, � (lif may be regarded as secondary, third-person accounts as upon the subjects the authors were trying__iQ__ex­
: ii of remote history, or they may be ( and indeed have pound_ Sometimes, indeed, they betray the author's
;.)
::1.:1:
been) 12 regarded as autobiographical writings of Gib­ perscm;lity, private thoughts, and social life more re­
bon, Droysen, and Taine_ Scholarly reviev,s of schol­ vealingly than they descfibe the things he had under
l,id,
iWi
iii!,
arly books ought to be among the least likely places to o15servation. Here, too, a document's significance may
. pf;:
;!11

• ·•.. ,I.
hunt for personal reactions ( except, as sometimes hap­
pens with ihe best reviews, the reviewer deliberately
have a greater relationship to the intention of the his­
torian than to that of the author_ S_ornetimes the his­
'··,,.g!,iii. sets out to present his own point of view); and yet torian may learn more about the author than the au­
'3'. ii'')''.
) LI!''

·-J.111\ifl how often private philosophies, attitudes, likes,. and thq.r intended that he should_ 13
,'., 1/,v·�!1li�

tfii�� n
�. dislikes are unintentionally betrayed by the most sober 13 Cf_ Allport, pp, 111-12, where the "unintentional personal

reviewers! Whether a document is to be examined for document" ·is discussed_

what it rev�als about its subject or for what it reveals

H
4. t-.
d:lli.e about its author - whether, in other words, it is a oJJ-- J.o-zµ,,.,.�� � � -��1:;:::__�'
J,eu.-� ,.Yu lct4r u,,0 < -r� ::. Q.t�
1\ �i1'1/1
;.! f. l(IJ:/
third-person or a first-person document - thus de­ 1 -.
. ;-r/J�!I
\ iii!! rl! :7.

-
pends upon the examiner's rather than the author's in­
tl,l,j i :·: · '-
/:;'· 11r111· � ·��
tention.
!§
E a (}
"". / ,,, l
• l!i[J
For the same reason the term personal document is

\�fl!
\ r.� �.-�
,.!!;(!'­
I

. to the historian synonymous with the term human �

<7 • �y � �t-t�3 -� ��
''
. ;1,·tf
)rl dccumeni. These terms were invented by social sci­
1"

s,i,rz.--.,f �
entists. The historian is not likely to employ them_ To
1 �:,
< •

.I,' � � �

1e\P.;.� -t-ti �a:.�

him they appear tautalogous. _b.11 documents are both �ul -�� �
�C2.u,)
!. ��
human and personal, since they are ilie-Rm:k__QLh_u­
<?�(1/4,w
� c�

tJ-L,__t \.u2-
:/!�·.filt,l
·.c111,;
jti!;t ti ri,;an beings and slied ligJ-it upon their authors as well 1ft;
.li ·-

12 Cf. J. W. Swain, "Edward Gibbon and the Decline of Rome,"

South Atlantic Quarterly, XXXIX (1940), 77-93; John R. Knipfing,
"German Historians and Macedonian Imperialisrh," American His­
.:n ll,i torical Re-view, XXXI (1921), 659-61; Louis Gottschalk, "French
· '1lo'i
i1t�,:ii! Revolution: Conspnacy or Circumstance" in Persecution and Lib­
erty,.Essays in Honor of George Lincoln Buu (New York, 1931), PP-
t;:
r_._
-:11Hl!i1
71 11•1 ,,. 445-52. Cf. J. H, Randall and George Haines, "Controlling Assump­
:{;jf tions in the Practice of American Historians," Merle Curti et a(.,
)\Jj!;; pp. 17-52, and H. K Beale, "What Historians Have Said about the
f!:1jj/
hi1r Causes of the Civil W2r," ibid., pp. 55-92,
r
.,:·;\;'\·1r:i1'i1
jfl!l'f/ i{:�\.
ill i l:i
•'· •/1, 'I
i; �

: jJ/1
.. ,
a

,,:if ;Ii I
�-. 111

I
 · :�;��{1t�.
[ 118 ] ./-" ·:n
Jjj·
THE PROBLEM· OF' AUTHENTICI

used to be cited on occasion to bols

TY

ter a theory that

m in the w est. In
119
'

the popes had a wide territorial c1ai

CHAPTER VI
� 1440 Lorenzo Valla proved, chie
fly by means of an­
f'
,, that it had been
THE PROBLEj\1 OF AUTHENTICITY, �'i achronisms of style and allusion,
l s are counterfeited
forged. At other times document
Marie Antoinette
OR EXTERNAL CRITICISM 4 for sale. Counterfeit letters of Queen
t�· used to tum up frequently. A Philad
2 elphia autograph
�i
nufactured hun­
So far it has been assumed that the documents dealt dealer named Robert Spring once ma
to supply the de­
with have been authentic. The problem of authenticity dreds of skillful fo rgeries in order
us example of
sel9om concerns the sociologist or psychologist or an­ mand of collectors. A recent notorio
Abraham Lincoln
thropologist, who generally has � living subject under �­
l forgery was the "correspondence" of
on the Atlantic
l and Ann Rutledge, palmed off
his eye, can see him as he prepares his autobiography,
anC:: can cross-examine him about doubtful points. Monthly in 1928.
J mercenary con-
Even in the law courts the question of authenticity of j Sometimes f abrication is d·ue to less
documents becomes a difficult problem only on rare l sidera tions. Political propaganda larg
el y accounts for
a "document" pre­
o..::cJ.sions, whe n the writer or witnesses to the writing the Protocols of the Elders of Zion,
conspiracy to rule
cannot be produced. 1 But for historical documents tending to reveal a ruthless Jewish
; are bas ed only
those occasions are not r are. They are in fact freque nt the world. Sometimes historical "facts"
3

of H. L. Menck­
for ma:c.uscript sources; and if doubt as to authenticity on some practical joke, as in the case
" of the bath­
arises less ofte� for printed sources, it is because us1.1- en's much cited article on the "history
cking letter of en­
ally some skilled editor h as already performed the task tub, or of Alexander Woollcott's mo
d ( of which he
of authenticating them. dorsement of Dorothy Parker's husban
pos ed addressee, al­
Forged o, Misleading Documents never sent the original to the sup
to the endorsee) .�
though he did send the carbon copy
are a striking ex-
Forgeries of documents in whole or in part ,,;,,ithout The Memoires of Madame d'Epinay
being usual, are common e�ough to keep the careful
ch Revolution (London,
histm-i:m consta ntly on his guard. "Historical docu­ 2 Lord Action, Lectures on the Fren
ments" are fabricated for several reasons. Sometim es 1910), pp. 361-4. the Protocols of Zion (New
s See J. S. Curtis, An Appraisal of
they a re used to bolster a false cla im or title. A well­ York, �942). 9;
k'Jown exampl e is the Donation of Constantine, which i Cf. C. D. MacDougall,
Hoaxes (New York, 1940), pp. 302-
t: His Life and His World
Dorothy Parker, reviewing A. Woo llcot
Sun Book Week
Wigmore, pp. 32Cr-36. York, 1945) in the Chic ago
1
by S. H. Adams (New
of June 10, 1945.

·.'.Ii
 ··
:-��
---THE-PROBI:;EM-0F-AU1'Hl':
NT1CIT.Y 121.
120
_(.li,
UNDERSTAN.DING HISTORy
<tt� Starrett to write his verse en­
kind that led Vincent
ample of fabricJtion of a whole book that has beguiled · .. \
ing for Fame":
'i titled "After Much Striv
f:il
even respectable historians.
5

think,
Sometimes quite genuine documents are intended It would be rather jolly, I
To be the original authority

i;·1'.
to mislead certain contemporaries and hence have
misled subsequent historians. A statement supposed I On some obscure matter of
literatur. e or faith
Upon which, in one's leisure,
.to. be that of Emperor Leopold II's views on the ccurate pamphlet;
French Re-,oluti on misled Marie Antoinette and sub­ . One had jotted down an ina
1
·:: And forever therea fter

-�l:
se quently even the m ost careful historians until it was centian borrowers
To be quoted by all post-Vin
exposed in 1894 as a wishful statement of s ome
In a pertinacious footnote.
! 9

French emigres. 6 In days when spies were expected to !

open mail in the post, write rs of letters would occa­ ations of the n ature of
Occasionally misrepresent It is still
sionally try to o utwit them by turning their curiosity editors' tricks.
printed works result from ­
� of the many writings attrib
to the advantage of the one spied upon rather than to a matter of dispute which
that of the spy or his employer. 7 And when censors ] uted to Cardinal Richelieu
were in fact written or
might condemn boo ks to be burned and writers to be ;j dictated by him; and little
of the so-calle d M
emoires
ment politiq-ue de Colbert
imprisoned, authors could hardly be blamed if they de Jean de Witt and Testa
hn de Witt and Colbert.
some.times signed others' names to their work. For in­ were in fact written by Jo
Condorcet and to Weber,
stance, it is hard to tell whether some works actually The memoirs attributed to
nette, and several works
written by Voltaire are not still ascribed to others. It foster-brother of Marie Antoi
by others than their al­
is thus p ossible to be too skeptical about a document ascribed to Napoleon I are ers have
of daily n ews pa p
which may be genuin e though not what it seems. leged authors. Even issues bear.
er the dates they
B ernheim has provided a list of documents that were been manufactured long aft
me good examples (see
once hypercritically considered unauthentic but are
The Moniteur furnishes so
now accepted." Perhaps it was hypercriticism · of this es of Napoleon have been
p. 107 above). Several Diari
writings. The circum­
made up by others from his tor­
ep resentation of his
stances of the forgery or misr
5 The "che:iting document"
is discussed witl: a wealth of absorb­
or­
ing detail in Allan Ne-,ins, Gateway to History ( Boston, 19 3 8), emselves r e veal imp
ch. V, pp. 119-37. ical documents m ay often th orm ati on -
graphical inf
6 Acton, French Revolution, p. 119.
tant political, cultural, and bio th y
s or persons a s if
e
but not "bout the same event
7 Cf. Lafayette to William Carmichael,
Maren 10, 1785, quoted
in Louis Gottschalk, Lafayette between the American and the
French Revolution (Chicago, 1950), pp. 156-7. .. were genuine.
8 Ernst Bernheim, Lehrbuch der
historischen Methode und der ett.
Geschichtsphilosophie ( 6tl1 ed.; Leipzig, 1908), pp. 376-9 1.
0 Quoted by permission of Mr. Starr
 TICITY.. _..:123.
122 · ·THE-PROBI,EM-OF--AUTHEN
UNDERSTANDING· HISTORY
sts who are familiar
Tests of Authenticity
}1]
"';•f'.31
:
tion) can be d etected by sp10eciali
with contemporary writing. Oft
en spelling, p articu­
.';Ji atures (because too
To distinguish a hoax or a misrepresentation from ,; ·f] larly of proper names and sign
), reveals a forgery,
a genuin e document, the historian has to use tests good or too bad or anachronistic
r. Anachronistic ref­
as would also unhistoric gramm a
: t:
that are common also iu police and legal .det ection. late or too remote}
erences to e-vents ( too early or too
Making the best guess he can of the date of the docu­ .. . .M : e when the al­
ment ( see below pp. 1 38 and 147-8), he examir es the .·: _ -,·,,
ll or the dating of a document at a tim
materials to s ee whether they are not anachronistic:
paper was rare :in Europe before the fifteenth century,
·1l leged writer could not possibly have
designated ( the alibi) uncovers frau
b een at the place
d. Sometim es the
and printing was unknown; pencils did not exist there l skillful forger has all too carefully
followed the b est
omes too obvi­
before the sixteenth c entury; typewriting was not in­ historical sources and his product bec
--� where, by skillful
v ented until the nineteenth crntury; and India paper j ously a copy in certain passages; or
ewd enough to
came only at the end of that century. T11e historian l paraphrase and invention, he is shr
is given away by
also examines the ink for signs of age or for anachro­ l
t avoid detection in that fashion, he
unknown details
nistic chemical composition. Making his best gu ess of the absence of tri-via and otherwise
ally, however,
from his manufactured account. 11 Usu
:j
the possible author of the document (see below pp. be - for ex­
to
144-7), he s ees :if he can identify the handwriting, if the document is where it ought
signature, seal, letterhead, o r watermark. Even when
the handwriting is unfamiliar, :it can be compared
1
-j
ampl e, in a family's archives, or amo
firm's or lawyer's papers, or in a gov
ng a business
ernmental bu­
it is in a li­
reau's records (but not merely b ec aus e
wich autl1enticated specimens. One of the unfulfill ed
ection) - its
n eeds of the historian is more of what the French call brary or in an amateur's autograph coll 12
wyers call it),
"isographies" - dictionari es of biography giving ex­ provenance ( or its custody, as the l a
ess.
amples of handwriting. For som e p eriods of history, creates a presumption of its genuinen
1 700,"
experts using techniques known as paleography :;.nd 1° Cf. Marcel Cohen, "Comment on parlait le fran�ais en
diplomatics, first systematized by Mabillon in th e L'Europe, XXV (1947 ), 18-2 3 .
11 Cf. They Knew the W ashingtons;
Letters from a French Sol-
sev enteenlh century ( see p. 12 7 b elow), h ave long ly in Virginia, tr. Princess
dier with Lafayette and from His Fami
kno_wn that in certain regions at certain tim es hand­ i Beraud, My Friend Ro•
Radziwill (Indianapolis, 1926); and Henr
writir.g and the style and form of official documents n (New York , 19 ).
bespierre, tr. Slater Brow
28

were more or less conventionalized. Seals have been 12 Wigmore,

P?· 330- -1 .
the subj ect of special study by sigillographers, and ex­ ·ji
p erts can detect fak ed ones ( see below, p. 128).
Anachronisiic c;tyle ( idiom, orthogr aphy, or punctua-
.I
-ij'
l
..,,,.
<:i
 124 U:.�DERSTANDING HISTORY
Garbled Documents
A document that in its entirety or in large part is
the result of a deliberate effort to d eceive may often
be hard to evaluate, but it sometimes causes less
trouble than does th e document that is unauthentic
only in small part. For such parts are usually the re­
s ult, not of studied falsehood, but of unintentional
ewJr. They occur r.1ost frequently in copies of docu­
ments whose originals have disappeared, and afe gen er­
ally dne to that kind of error of-omission, repetition,
'Jr addition with which anyone who has ev er made
copies �oon becomes familiar. Sometimes they are the
result, h0wever, not of carelessness but of deliberate
intention to modify, supplement; or continue the
original. Such a cbange may b e made in good faith in
H..e first instance, care being exerted to indicate the
differences b etween the original text and th e glossary .
,Jr continuations, but future copyists are often less
careful or more confused and make no such distinc­
tions.
This p roblem is most familiar to classical philolo­
gists and Bible critics . For th ey seldom have copies
less than eight centuries and s everal stages of repro­
duct:on removed from the original - that is to say,
copies of copies of copi es, and sometimes copies of
translations of copies of translations of copies, and so
on. The philologists give to this problem of estab­
lishing an accurate text the name textue1l critici.sm,
and in Biblical studies it is also called lower criticism.
The historian has borrowed his t echnique from phi-
1ohgists and Bible cri tics .
THE PROBLEM OF-AUTHENTICITY- -125·
The Restoration of Texts
a,
� The technique is complicat ed but can be briefly
r
Ii
.1 ·
described. The first task is to collect as many copies of
the dubious text as diligent search will reveal. Then
they are compared. It is found that some contain
I
t}';
words or phrases or whole passages that are not con-
. tained in others. The question then arises: Are those
words, phrases or passages additi�ns to the original ·
text that have found their way into some copies, or are
4
J
they omissions from the others? To answer that ques­
f
; tion it is n ecessary to divide the available copies into
?:: one or more "families" - that is, groups of texts
;
which closely resemble each other and therefore seem
to be derived, directly or indirectly, from the same
master copy. Then by a comparison of the t exts within
each family an effort is made to establish the com­
parative age of each in relation to the others . If the
members of the same family are largely copied from
each other, as this arrangement in families frequently
shows, the oldest one is in all probability (but not
necessarily) the one near est the original. This ,process
j
.I
is continued for all the families. When the copy near­
··
;,,�,
1 est the original in each family is discovered, a com­
parison of all of t hese "father" copies will usually
I
.;
I
then reveal words and passages that are in some but
not in others. Again the question arises: Are those
words and passages additions to the copies that have
them or omission s from the copies that do not? The
most accurate available wo�dings of the passages
added or omitted by the respective copyists are then
prepar ed. Changes in handwritings, anachronisms in
 126 tJNDERSTANDING· HISTORY
style, grammar, orthography, or factual detail, and
opinions or errors unlikely to have been those of the
original author frequently reveal additions by a later
hand. When the style and contents of passages under
..J,
•\. •:"
discussion may be attributed to the author, it is safe
to assurr,e that they were parts of his original manu­
script but were omitted by later copyists; and when
they cannot be attributed to the author, it is safe to
assume that they were not parts of his original manu­
script. In some casc:s, a final decision has to await the
discovery of still more copies. In ·many instances the
original text can be approximately or entirely restored.
By a �imilar method one can even guess the· con­
tents, at least in part, of a. "father" manuscript even
when no full copy of it is in existence. The historian
Wilhelm von Giesebrecht, a student of Ranke, at­
tempted to reconstruct a text that he reasoned must
be the ancestor of several eleventh-century chronicles
in which he had noted striking similarities. By adding
together the passages that appeared to be "descended"
from an unknown chronicle, he made a guess as to its
contents. Over a quarter of a century later the ancestor
chronicle was in fact found, and proved to be exten­
s:vely like Giesebrecht's guess.
Sciences Auxiliary to History
The problem of textual restoration does not fre­
quently disturb thP, present-day historian, chiefly be­
ca,.1se many experts, engaged in what the historian
egocentrically calls "sciences auxiliary to history,"
provide him with critically prepared texts. Since
J P,an Frarn;:ois Champollion in 1822 learned to de-
THE PROBLEM OF AUTHENTICITY 121·
cipher hieroglyphics, part of the work of Egyptologists
and papyrologists has been to provide the historian
with texts and translations of inscriptions and papyri
r found in the ancient Nile Valley, whether in Egyp­
l tian hieroglyphic or in cursive hieratic and demotic or
in Greek. The Assyriologists, since Sir Henry Rawlin­
son in 1847 deciphered Old Persian cuneiform and
in 1850 Babylonian cuneiform, have been publishir.g
and translating the texts found on the clay tablets of
the ancient Tigris-Euphrates civilizations. Biblical
J criticism, even before Erasmus, was directed to the
l
effort of bringing the text of both the Old and the
,.j= New Testament as close as possible to the original
1 wording and of explaining as fully as possible the
1 Hebrew and Hellenistic civilizations which they re­
flected. Philology, as already explained, deals among
other things with the derivation from variant texts of
the most authentic ones ( especially of classical liter­
ature). The classical epigrapher restores and edits the
texts of Greek �md Latin inscriptions found on the
gravestones, monuments, and buildings of ancient
Greece and Rome. The paleographer, since the time
that Mabillon (see p. 122) first formalized the prin­
ciples of paleography and diplomatics, has been able
to authenticate medieval charters and other docu­
ments by their handwritings, which have been found
to vary from place to place and from time to time, and
by their variant but highly stylized conventions and
forms, and to publish easily legible printed versions of
them. The archeologist excavates ancient sites and
provides the historian with information derived from
artifacts such as statues, mausoleums, pottery, build-
 138 UNDERSTANDING-HISTORY
tional form to motives he had imperfectly analyzed,
that she laid bare features in his character he had
never realized." 11 If Morris R. Cohen is right, "To
widen our ho:izon, to make us see other points of
view than those to which we are accustomed, is the
greatest service that can be rendered by the historian,
and this he can do best by concentrating on the spe­
cial field which he studies to understand." 18
Identification of Author and of Date
Some guess cf the approximate- date of the docu­
ment and some identification of its supposed author
( or, at least, a surmise as to his location in time and
space and as to his habits, attitudes, character, learn­
ing, associates, etc.) obviously form an essential part
of external crdicism. Otherwise it would be impossible
to prove or disprove authenticity by anachronisms,
handwriting, style, alibi, or other tests that are as­
sociated with the author's milieu, personality, and
actions. But similar knowledge or guesses are also nec­
essary for internal criticism, and therefore the prob­
lem ol author-identification has been left for the next
chapter (pp. 144-8).
Having established an authentic text and discov­
ered what its author really intended to say, the hi:;­
torian has only established what the witness's testi­
mony is. He h::;.s yet to determine whether that
testimony is at all credible, and if so, to what extent.
That is the problem of internal criticism.
17 January 21, 1881, Herbert Pau! (ed.), Letters of Lord Acton
to Mary Gladstone (New York, 1904), p. 159.
18 The Meaning of Human History (La Salle, Ill., 1947), p. 28.
- [ 139 ··]
,,E'-f
:l
� CHAPTER VII
�
;I"
" THE ·PROBLEM OF CREDIBILITY, ·
ii
:I
OR INTERNAL CRITICISM
i'
r
THE HISTORlAN first aims in the examination of tes­
timony to obtain a set of particulars relevant to some
topic or question that he has in mind. Isolated par­
ticulars have little meaning by themselves, and unless
they have a context or fit into a hypothesis they are of
doubtful value. But that is a problem of synthesis,
which will be discussed later. 1 What we are now
concerned with is the analysis of documents for cred­
ible details to be fitted into a hypothesis or context.
What Is Historical Fact?
In the process of analysis the historian should con­
stantly keep in mind the relevant particulars within
the document rather than the document as a whole.
i.
r!'
Regarding each particular he asks: Is it credible? It
might be well to point out again that wh2t is meant
i: by calling a pa1ticular credible is not that it is actt!ally
I
what happened, but that it is as close to what actually
happened as we can learn from a critical examination
of the best available sources. 2 This means verisimilar
at a high level. It connotes something more than
merely not being preposterous in itself or even than
ll
plausible and yet is· short of meaning accurately de-
n-
1 See Chapter X.
.J
2 Cf. above pp. 45-<}·
 '1�
140 UNDERSTANDING HISTORY THE PROBLEM OF CREDIBILITY 141
scriptive of past actuality. In other words, the histor­ not self-evident, like the Pantheon and Chinese.litera­
ian establishes verisimilitude rather than objective ture), involve no judgments of value ( except with
tr"'.1.th. Though there is a high correlation between the regard to the antiquity of Chinese literature), contra­
two, they are not necessarily identical. As far as mere L4, dict no other knowledge available to us, seem other­
partic 1lars are concerned, historians disagree relatively .£tali
,, i1,:f wise logically acceptable, and, avoiding generalization,
seldom regarding what is credible in this special sense •·t1 J
f11•·. deal with single instances.
of "conforming to a critical examination of the
sources." It is not inconceivable that, in dealing with ·rql·t·,,(!,
Even some apparently simple and concrete state­
ments, however, are subject to question. If no one
the same document, two historians of equal ability i
disputes the historicity of Socrates, there is less agree­
,ft
and training would extract the same isolated "facts" •ti
.-:!.;···r!•'.'
ment regarding Moses and earlier figures of Hebrew
.md agree with each other's findings. In that way the
elementary data cf history are subject to proof.
,:
f folklore. If no one doubts that Michelangelo sculptured
his "Moses," a few still think that Shakespeare's plays
:··.'fg-,_:.:i ..
�i-_..'

A historic:il "fact" thus may be defined as a partic­ were in fact written by Francis Bacon. Doubt regard­
ular derived directly or indirectly from historical docu­ ing concrete particulars is likely to be due, however,
ments and regarded as credible after careful testing r,l·'
fl· to lack of testimony based on first-hand observation

ft;
in accordance with the canons of historical method
(see below p. 150). An infinity and a multiple variety
of fac-ts of this kind are accepted by all historians:
e.g., that Socrates really existed; that Alexander in­
vaded India; that the Romans built the Pantheon;
that the Chinese have an ancient literature (but here
we introduce a complexity with the word ancient,
ll.:
!.I
rather than to disagreement among the witnesses. In
general, on simple and concrete matters where testi­
mony of direct observation is available, the testimony
can usually be submitted to tests of reliability that
will be convincing either pro or con to most competent
and impartial historians. As soon as abstractions, value
judgments, generalizations, and other complexities en­
. .

;j_, '
which needs definition before its factual quality can ter into testimony the possibility of contradition and
be considered certain); that Pope Innocent III ex­ { '.j debate enters with them. Hence, alongside the mul­
communicated King John of England; that Michel­
i·,
I ·\
titude of facts generally accepted by historians, exists
angelo sculptured "Moses"; that Bismarck modified another multitude debated ( or at least debatable) by
the dispatch from Ems of King William's secretary; them.

t·'f
i·"•:_. ;
that banks in the United States in 1933 were closed
for four days by presidential proclamation; and that T'!,e Interrogative Hypothesis
"the Yankees" won the "World Series" in 1949. Sim­ In analyzing a document for its isolated "facts,"
ple and fully attested "facts" of this kind are rarely {1·:i :
the historian should approach it with a question or
di:;putcd. They are easily observed, easily recorded ( if f.
J
:t⇒ � ·,. a set of questions in mind. 111e questions may be rela-
hm'' ;
. f'!i.t1: t\
-�- {'

'tt·�\
 · 143
142 UNDERSTAND_�9 HISTORY THE PROBLEM OF CREDIBILITY

il
tively noncommittal. (E.g.: Did Saul try to assassi­ low), since only those materials are relevant ·which
nate David? What were the details of Catiline's life? lead directly to an answer to the question or indicate
Who werl'! the crusading companions of Tancred? that there is no satisfactory answer.

What was the date of Erasmus' birth? How many
men were aboard De Grasse's fleet in 178 1 ? What is
tlie correct spelling of Sjeyes? "\Vas Hung Hsui-chu'an
a Christian?) Jt will be noted that one cannot ask
e-ven simple questions like these without knowing
enough about rnme problem in history to ask a ques­
tion about it, and if one knows enough to ask even
the simplest question, one already has some idea and
probably some hypothesis regarding it, whether im­
plicit or explicit, whether tentative and flexible or
fo1mulated and fixed. Or the hypothesis may be full­
hf
l. F.'·':'
The Quest for Particular Details of Testimony
As has already been pointed out, every historical
�!.:�- subject has four aspects - the biographical, the geo­
graphical, the chronological, and the occupational or
functional. With a set of names, dates, and key-words
: ifr in mind for each of these aspects, the historical in­
vestigator combs his document for relevant particulars
f.•
(or "notes," as he is more likely to call them). It is
','[J. ,. generally wise to take notes on relevant matter
/�. •i1·:�1-,. whether or not it at first appears credible. It may turn
out that even false or mistaken testimony has rele­

J,,j
fledged, though still implicit and in interrogative
form. (E.g.: Can the Jews be held responsible for . } '�-!;1_ vance to an understanding of one's problem.
: n, :: t
the crucifixion of Jesus? Did the medieval city de­ Having accumulated his notes, the investigator
ti;l
C
velop from the fair? V/hy did the Anabaptists bel ieve must now separate the credible from the incredible.
Even from his "notes" he has sometimes to extract

/;:i:�: tf( j
in religious l iberty? How did participation in the
American Revolution contribute to the spread of lib­ still smaller details, for even a single name may reveal

·:,:<
eral ideas among the French aristocracy? Why did a companion of Tancred, a single letter the correct
Woodrow Wilson deny knowledge of the "secret J ', •... _.._
spelling of Sieyes, a single digit the exact number of
De Grasse's crew, or a single phrase the motives of

· i/·1'\:ti'.'
treaties"?) In each of these questions a certain im­
p lication is assumed to be true and further clarifica­ Wilson's denial. In detailed investigations few docu­
-. · �,f; [1I ments are significant as a whole; they serve most often
tion of it is sought on an additional working assump­
tion. only as mines from which to extract historical ore.
ti- (g_.
Putting the hypothesis in interrogative form is more Each bit of ore, however, may contain flaws of its

.·:,tfflr :;
own. The general reliability of an author, in other
judicious tha!l putting it in declarative form if for
·.·• r_:;_
. .
;,
no other reason than that it is more noncommittal words, has significance only as establishing the prob­

,n<·•:
df
before all the evidence has been examined. It may also able credibility of his particular statements. From
help in some small way to solve the delicate problem that process of scrupulous analysis emerges_ an im­
of relevance of subject matter ( see Chapter X be- portant general rule: for each p articular of a docu-
,.,l'·1·t1· •l
It I

_;I:�
t)i'-�-}:

L���-. i\1
�t. - �b-�--
�--- ' t·"i'

.. ,,_I 11:
··.�.:.:., ..... ,•
 :i_'i;r·fl{;:
'',!,.�
-�·�

f: THE PROB1:'EM OF CREDIBILITY 145.

144 UNDERSTANDING HISTORY '
J of course, contain explicit biographical det ails, but
ment the process of establishing credibility should be ..- I to assume that would be begging the question. Even
separately undertaken regardless of the general credi­ ·i
bility of the author. where it is relatively free from first-person allusions,
r
much may be learned of the author's mental proc­
Identification of Author esses and personal attitudes from it alone.
s me Let us take the usual text of Lincoln's Gettysburg
As has already been pointed out (p. 1 38), o
t a doc­ Address, and assume for the sake of example that we
identification of the author is necessa1y to tes -�
eque nt proc ess of J have no knowledge of it except for what its own con­
ument's authen ticity. ·In the subs .\
s, even the tents may reveal:
determining the credibility of its particular . ·I:
as
most genuine of documents sh�uld be regarded l
"/, Fourscore and seven years ago our fathers brought forth
inno cent . Tbe imp or ­
guilty of deceit until proven ;·.· \

! on this continent a new nation, conceived in liberty, and

's gene ral_ reli­
tance of first establishing the author i
,/ dedicated to the proposition that all men are created
the
ability is therefcre obvious. Where the name of equal. Now we are engaged in a great civil war, testing
t
author can be determined and he is a person abou
ificat ion ·.',J whether that nation, or any nation so conceived and so

whom biographical data are avail able, ident
is z. relatively ea�y task. Because, in mos t legal and
social science in-1estigations, the wit ness or the autho
n and availa ble
of a document is personally know
t
the investigator, that question generally presents no
\
; dedicated, can long endure. vVe are met on a great battle­
--i field of that war. We have come to dedicate a portion of
r
that field, as a final resting-place for those who here gave
o their lives that that nation might live. It is altogether fit­
ting a�d proper that we should do this. But, in a larger

r,l1
sense, we cannot dedicate - we cannot consecrate - we
insurmountable difficulties to lawyers and social sci- cannot hallow - this ground. The brave men, living and
er.tists.

:i
dead, who struggled here, have consecrated it, far above
The historian, however, is frequently obliged to use our poor power to add or detract. The world will little
... d 0 cuments written by persons about whom nothing or note, nor long remember, what we say here, but it can

r
relatively little is knov.rn. Even the hundreds· of bio­
graphical dictionaries a11d encyclopedias already in
existence may he of no help because the a1-1thor's
i
f name is unl-nown or, if known, not to be found in
t he re:-erence work s. The historian
must therefore de­
pend upon the document itself t o teach him what it
can about the author. A single brief document may
questions. It may,
t each him much if he asks the right
never forget what they did here. It is for us the living,
"" rather, to be dedicated here to the unfinished work which
they who fought here have thus far so nobly advanced. It
-t is rather for us to be here dedicated to the great task
remaining before us, - that from these honored dead we
take increased devotion to that cause for which they gave
the last full measure of devotion - that we here highly re­
solve that these dead shall not have died in vain - that

!
-�,)�
·,:- ,,
•' i.
I:;,

;';\�. f
--_::�:,
 �·tr¥r�
146 UNDERSTANDING HISTORY
this nation, ur,der God, shall have a new birth of freedom
- and that gov.ernmcnt of the people, by the people, for
the people, shall J10t perish from the earth.
Even a hasty examination will suffice to make clea1
that the author, at the time of writing, was planning
. �
to use 1t·as a speech("we are met,""w hat we say
here"), that he wrote English we11, that his address
was a funeral oration("we have come to dedicate a
portion of that field as a final resting place"), that
he was probably a prominent citizen, that he pre­
sumably was an American("our fathers,""this con­
tinent,""new nation," "four score and seven ·years
ago"), that he was an advocate of liberty and equality
( or at least desired his hearers to think so), that he
lived during the American Civil War, that he was
sreaking at Gettysburg, or possibly Vicks burg("great
battlefield,""four score and seven years ago"), aad
that he wanted his side in the war to be thought of as
fighting for democracy( "government of the people,
by the people, for the people"). If we forget the con­
troversy among historians as to whether the words
under God were actually delivered or were only after­
ward inserted, we may assume that he subscribed, or
wished to appear to subscr:ibe, to the belief in a Su­
preme Being.
From a short document, it would thus appear, it is
possible to learn much about the author without
kncwing who he was. In the case of the Gettysburg
Address a trained historian would probc:bly soon de­
tect Li.ncoln's authorship, if it were unknown. But
even if he had never heard of Lincoln, he would be
i
�­ T-HE PROBLEM OF CREDIBILITY 147
1
\
I
!
able to tell that, in attempting to judge the .truth
of the particulars stated in that address, he would have
to consider it as probably a public exhortation by a
prominent antislavery Northerner after a major vic­
tory over the Confederate States in the American
Civil War. Many documents, being less modest and
less economical of words than the Gettysburg Ad­
fl dress, give their authors away more readily.
Determination of Appro�imate Date
It would be relatively easy, even if the Gettysburg
Address were a totally strange document, to establish
its approximate date. It was obviously composed
"four-score and seven years" after the Declaration of
Independence, hence in 1863. But few strange docu­
ments are so easily dated. One has frequently to re­
'( sort to the conjectures known to the historian as the
terminus non ante quern ("the point not before
which") and the terminus non post quern "the( point
not after which"). These termini, or points, have to
be established by internal evidence - by clues given
within the do_cument itself. If the date 1863 were not
implicit in the Gettysburg Address, other references
within the speech could point obviously to the be­
·�1· ginning of the American Civil War as its terminus
non ante quern, and since the war was obviously still
going on when the document was composed, its ter­
minus non post quern would be the end of the Civil
\tVar. Hence its dat� could be fixed approxiIT! ate]y,
even if the first sentence had been lost, as somewhere
between 186i and 1865; and if we were enabled by
i'
� other data to guess at "the great batt]efield," we

.,- �l
148 UNDERSTA.!'l'DING HISTORY
might even narrow that margin. Some documents
·.E
might not permit even a remote guess of their ter­
mini, but where the author is known, one has at
least the dates of his birth and death to go by.
The Personal Equation
This analysis of the Gettysburg Address ( under
the false assumption that its authorship is unknown)
in :iicates the type of question the historian asks of
both anonymous and avowed documents. Was the
author an eyewitness of the events.he narrates? If not,
what were his sources of information? When did he
write the document? How much time elapsed be­
tween the event and the record? What was his pur­
pose in writing or speaking? Who were his audience
and why? Such questions enable the historian to an­
swer the still more important questions: vVas the
author of the document able to tell the truth; and if
able, was he willing to do so? The ability and the
willingness of a witness to give dependable testimony
are determined by a number of factors in his per­
sonality and social situation that together are some­
times called his "personal equation," a term applied
to the correction required in astronomical observa­
tions to allow for the habitual inaccuracy of individ­
ual observers. The personal equation of a historian is
sometimes also called his "frame of reference," but it
probably will be found more expedient to restrict the
latter term to his conscious philosophy or philoso­
phies of life in so far as they can be divorced from
personality traits and biases of which he· · may or may
not be aware.
,I
..
,,,,
! >
. ij
• THE PROBLEM OF CREDIBILITY
;
149
General Rules
In a law court it is frequently assumed that all the
testimony of a witness, though under oath, is suspect
\'
if the opposing lawyers can impugn his general char­
acter or by examination and cross-examination create
doubt of his veracity in some regard. Even in modem
law courts the old maxim falsus in uno, falsus in om­
nibus tends to be overemphasized. 3 In addition, hear­
say evidence is as a general rule excluded; 4 certain
kinds of witnesses are "privileged" or "unqualified"
and therefore are not obliged to testify or are kept
from testifying; s and evidence obtained by certain
means regarded as transgressing the citizen's rights -
such as "third degree," drugs, wire-tapping, or lie­
detector - are ruled out of some courts. The legal
1· system of evidence, says James Bradley Thayer, "is not
concerned with nice definitions, or the exacter aca­
demic operations of the logical faculty.... Its rules
... are seeking to determine, not what is or is not,
in its nature, probative, but rather, passing by that
inquiry, what among really probative matters, shall,
nevertheless, for this or that practical reason, be ex­
cluded, and not even heard by the jury." 6 Courts of
law, in the Anglo-Saxon system at least, go on the
·i
assumption that if one side presents all the permis­
sible testimony in its favor and if the other side pre­
sents all the permissible testimony in its, the truth
Wigmore, p. 181.
3
Ibid., pp. 238-45.
4
0
Ibid., pp. 125-34 and 354-60.
l
6 Preliminary
Treatise on Evidence at thfl Common Law
(Bo$�On, 1896), pp. 3-1-
•

,i
150 UNDERSTANDING HISTORY
will emerge plainly enough fer judge and jury from
the conflic-:: or harmony of the testimony, evP,n if some
kinds of testimony are not permissible; and possibly
where much and recent testimony is avail2ble, the
innocent suffer less often by such an assumption than
the guiity escape.
The historian, however, is prosecutor, attorney for
the defense, judge, and jury all in one. But as judge
he rules out no evidence whatever if it is relevant.
To him any single detail of testimony is credible·­
even if it is contained in a document obtained by
force or fraud, or is otherwise impeachable, or is
based on hearsay evidence, or is from an interested
witness - provided it can pas:; four tests:
( 1) Was the ultimate source of the detail (the
primary witness) able to tell the truth?
( 2) Was the primary witness willing to tell the
truth?
( 3) Is the primary witness accurately reported with
regard to the detail under examination?
(4) Is there any independent corroboration of the
detail under examination?
Any detail (regardless of what the source or ·who the
author) that passes all four tests is credible historical
evidence. It wi11 bear repetition that the primary
witness and the detail are now the subjects of exam­
ination, not the source as a whole.
Ability to Tell the Truth
( 1) Ability to tell the truth rests in part upon the
witness's nearness to the event. Nearness is here used
in both a geographical and a chronological sense. The
r
..;,1
:1
i THE PROBLEM OF CREDIBILITY
i
\' reliability of the witness's testimony tends to va:ry
in proportion to (a) his own remoteness from the
scene in time and space, and (b) the remoteness
from the event in time and space of his recording of
it. There are three steps in historical testimony: ob­
servation, recollection, and recording (not to men­
tion the historian's own perception of the witness's
record). At each of these steps something of the pos­
sible testimony may be lost. G�ographical as well as
chronological closeness to the event affects all three
steps and helps to determine both how much will. be
lost and the accuracy of what is retained.
( 2) Obviously all witnesses even if equally close to
the event are not equally competent as witnesses.
Competence depends upon degree of expertness, state
of mental and physical health, age, education, mem­
T ory, narrative skill, etc. The ability to estimate num­
bers is especially subject to suspicion. The size of the
army with which Xerxes invaded Greece in 480 B.C.
was said by Herodotus to have numbered 1,700,000,
but it can be shown to have been considerably less
by the simple computation of the length of time it
would have taken that many men to march through
the Thermopylae Pass even unopposed. More re­
cently by a similar computation doubt was thrown
upon the veracity of a newspaper report from Moscow
that one million men, women, and children paraded
through the Red Square in celebration of the thirty­
second ar;.miversary of the October Revolution (No­
vember 7, 1949) ir. a five and one-half hour demon­
stration, for it would require more than fifty persons
a second to march abreast past a given point to com-
 152 UNDERSTANDING HISTORY
plete a parade of one million in five and one-half
hours. 1 With some notable exceptions, such as the
Domesday Book of William the Conqueror, histor­
ians have been war:!.1ed against using any source of
numbers before the end of the Middle Ages. 8 The
careful keeping of vital statistics was a relatively late
innovation of the end of the eighteenth and the be­
ginning of the nineteenth century. Previous to that
time tax rolls and incomplete parish records of bap­
tisms, marriages, and burials were the best in­
dications. Even battle casualty statistics before the
nineteenth century are suspect, and historians still
&;agree on the cost in human life of wars up to· and
including those of Napoleon I, and, in some instances,
beyond.
( 3) Degree of attention is also an important factor
in the ability to tell the truth. A well-known story, no
less illustrative if it be apocryphal, tells of a psychology
professor who deliberately staged a fight in his class­
room between two students, which led to a free-.for-all.
When peace was restored, the professor asked each
r.iember of the class to write an account of what had
happered. There were, of course, conflicting state­
ments among the accounts, but, what was most sig­
nificant, no student had noticed that the professor in
the midst of the pandemonium had taken out a
banana and had peeled and eaten it. Obviously the
entire meaning of the event rested upon the unno-
7 Letter of John E. Frazer, November 9, 1949, New York Times,
'rfovember 15, 1949.
8 Seignobos, 1vfethode historique appliquee aux sciences sociales,
pp. 204-5.
THE PROBLEM OF CF.EDIBILITY 153
ticed act; it was an experiment in the psychology of
attention. Because each student's interest had· been
fixed upon his own part in the drama, each had given
an erroneous interpretation of what had occurred.
Magicians similarly depend upon their ability to divert
attention from things they are doing to perpetrate
some of their tricks. The common human inability to
see things clearly and whole makes even the best of
witnesses suspect.
(4) We have already discussed the danger of the
leading question (p. 104). Such questions, by imply­
ing the expected answer, make it difficult to tell the
whole truth. Lawyers also count the hypothetical
question ("Supposing you did agree with me, would
you act as I?"), and the argumentative or "loaded"
question ("Have you stopped beating your wife?"),
�!r and the coached answer as belonging to kindred cate­
I
gories. 9 Such questions are especially liable to be mis­
leading if they have to be answered "Yes" or "No."
Allport gives a striking illustration of the kind of mis­
information that can be derived from the witness_
whose narrative is circumscribed by the questioner.
He mentions an investigator who "secured fifty top­
ical autobiographies, forcing all writers to tell about
radicalism and conservatism in their lives," and who
j.,
from those biographies almost (but fortunately not
quite) came to the conclusion that "radicalism­
conservatism constitutes one of those first-order vari­
ables of which all personalities are compounded." 10
( 5) In the last instance the investigator barely ·
9 Cf. Wigmore, pp. 147-50 and 160-2.
10 Allport, p. 137.
 154 UNDERSTANDING HISTORY
missed reasoning in a circle - from premise back to
premise again. It has been contended also that one of
the reasons why religious problems and events receive
so much attention in the history of the Middle Ages
is that its pr1ncipal sources were written by clergy­
men. If medieval architects, landowners, soldiers, or
merchants had written more, they might have asked
and answered different kinds of questions and given
a different picture of medieval life. Possibly, if the
writmgs of our own intellectuals should prove to be
the major source for future accounts oi our age, future
hdorians will be �isled into thinking that intellectu­
als bad a greater influence upon human affairs iri our
time than they actually have. This sort of circular
argument must be especially guarded against when
an effort is being made to ascribe unsigned writings
to a supposed author, for it is easy to assume that the
ideas of the writings are characteristic of the supposed
autbor if those very articles a::e the basis of the as­
sumptions regarding the author's characteristics.
(6) One almost inescapable shortcoming of the
personal docur.1ent is its egocentrism. It is to be ex­
pected that even a modest observer will tell what he
himself heard and what he himself did as if those de­
tails were the most important things that were said
and done. Often it is impossible for him to tell his
story in any oth�:: terms, since that is the only way
he knows it. This observation is a more or less in­
evi':able corollary of the caution with regard to atten­
tion discussed above. · The famous speech of the
Ccmte de Mirabeau after Louis XVI's Royal Session
of June 2 3, 1 789, provides a pat illustration of how
THE PROBLEM OF CREDIBILITY 155
easily such egocentrism may mislead the historian.
Mirabeau ( though speaking in the third person) told
how he had said something about the necessity of
force: "For we shall leave our seats only by the power
of the bayonet." He failed to mention that several
I others were expressing a similar determination at
}' about the same time, though probably in more mod­
erate language. Therefore historians trusting too con­
·f
fidently to Mirabeau's testimony have sometimes
made him the heroic center of a desperate crisis; still
it is more probable that he was not so compicuous or
the situation so dramatic as he implied. 11
In general, inability to tell the truth leads to errors
of omission, rather than commission, because of lack
/i of completeness or lack o_f balance in observation,
recollection, or narrative. Such errors may give a pic­
f ture that is out of perspective because it subordinates
or fails to in-:lude some important things and over­
emphasizes those it does include.
Willingness to Tell the Truth
The historian also has to deal with documents
'I whose authors, though otherwise competent to tell
'') ':� the truth, consciously or unconsciously tell falsehoods.
There are several conditions that tend especially to­
\' ward untruthfulness i:fnd against which the experience
t
of mankind has armed lawyers, historians, and others
who deal with testimony. 12
11 Cf. F. M. and H. D., Fling, Source Problems of the French
Revolution (New York, 1913), p. 129; cf. also pp. 123, 139; 144,
i
/1
and 148.
1� Wigmore, pp. 1 76--(;o.
i
 156 UNDERSTANDING HISTORY
( 1) One of the most elementary rules in the anal­
ysis of testimony is that which requires the exercise
of caution :i.gainst the interested witness. A witness's
interest.is obvious when he himself may benefit from
perversiqn of the truth or may thereby benefit some
one or some cause dear to him. Certain kinds of propa­
ganda are perhaps the worst examples of deliberate
perversion of truth out of a desire to benefit a cause.
In the seventeenth century the word propaganda was
applied to Catholi.c missionary work without dis­
paragement. Since the nincteenth-·century, however,
it has been used more or less derogatorily to designate
any kind of concerted movement to persuade and ·the
instruments of such persuasion. The word may be
modern, but propaganda and its methods have been
familiar since efforts were first made to influe!1ce pub­
lic opinion.
( 2) Often the benefit to be derived from a perver­
sion of the truth is subtle and may not be realized by
the witness himself. In such a case the cause of pre­
varication probably is bias. If the witness's bias is favor­
able to the subject of his testimony, it is frequently
designated studium. If it is unfavorable, it may be
designated odium or ira. The Latin words are de­
rived from a declaiation by Tacitus that he would
write history sine ira et studio (thereby setting a
standard that few historians, including Tacitus, have
been able to achieve). Studium and odium, bias for
and bias against, frequently depend upon the wit­
ness's social circnmstances and may operate in a fash­
ion of which he himself may not be 2waie. It be­
come: important to the historian to know what the
THE PROBLEM OF CREDIBILITY 157
witness's Weltanschauung ( or "frame of reference")
may be, as well as his religious, political, social, eco­
nomic, racial, national, regional, local, family, per­
sonal, and other ties (or "personal equation"). Any of
these factors may dictate a predilection or a prejudice
that will shade his testimony with nuances that other­
,J)
wise might have been absent.
( 3) The intended hearers or readers of a document,
it has already been remarked (p. 90), play an im­
portant part in determining the truthfulness of a
statement. The desire to please or to displease may
lead to the coloring or the avoidance of the truth.
Speakers at political rallies and at banquets, writers
of wartime dispatches and communiques, makers of
-;.
I
polite letters and conversation are among the numer­
ous producers of documents that may subtly pervert
l fact for that reason. Akin to and often associated with
interest and bias, which are often socially determined,
this motive is nevertheless different from them, being
usually personal and individual. It may occasionally
I stand alone as an explanation of prevarication.
1· (4) Literary style sometimes dictates the sacrifice
I
l
of truth. Epigrams and - notoriously - slogans of
�!t,i
war and politics ( "L' etat c'est moi"; "Millions for de­
fense but not one cent for tribute"; "The Old Guard
'I'
dies but never surrenders"), 13 if properly discredited
in the interests of accuracy and truthful reporting,
would be robbed of pithiness and color. Authors of
autobiographies and letters, especially when they write
13 Paul Harsin, Comment on ecrit l'histoire (Paris, 1933; Eng­
i. lish :idaptation, Berkeley, 1935) contains an appendix that cites
-l a number of other examples with critical analysis of their origin.
'r

158 UNDERSTANDING HISTORY
for private amuser:i.ent, may feel tempted to state as
fact what is only hearsay or tradition or even fiction;
and frequently narrator� and reporters ( especially if
they hope for large audiences) seek to appear omnis­
cient rather than to use the less vigorous word, the
less striking phrase, the ifs and buts, the there-is-some­
reason-to-belie-ve and the it-is-perhaps-safe-to-say of
more precise discourse..
The anecdote is especially suspect. Much too often
it is a subsequent invention to throw into humorous or
striking relief some spectacular figure or episode. The
more apposite the anecdote, the more dubious it is
likely to be without corroboration. And yet the ex­
istence of an especially pat anecdote has a historical
significance of its own - as showing the sort of thing
believed of or imputed to the subject. A well-known
Italian proverb describes such anecdotes as felicitous
(ben trovato) even if untrue.
( 5) Laws and conventions sometimes oblige wit­
nesses to depart from strict veracity. The same laws of
libel and of good taste that have encouraged the hid­
ing of the "resemblance to persons now living or
dead" in fiction and moving pictures have precluded
complete accuracy in some works of history. Some
of the notcrious inaccuracies of Jared Sparks as a his­
torian were due to his writing of living characters
from testimony by living witnesses who requested
him not to use certain data. 14 Etiquette in letters
and conversation, conventions and formalities in treat­
ies and public documents require politeness and ex-
u Louis Gottschalk, Lafayette and the Close of the American
Revolution (Chicaga, 1942), p. 252.
I,
r
THE PROBLEM OF CREDIBILITY 159
j
pressions · of esteem that are obviously false or empty.
A successful comedy, James Montgomery's Nothing
but the Truth (1916), was written around the valiant
effort of a young man to go through a whole day with­
out saying anything that was untrue; it nearly cost
I him all his friends. Religious concepts like the Chris­
=p
! tian Scientist's interpretation of the ideas of evil, dis­
'r ease, and death may lead to misunderstanding. Cor­
porations, commissions, and societies are sometimes
required by their articles of incorporation or constitu­
tions to meet periodically, but when their numbers
are small, the minutes of their meetings may be much
more formal than the actual meetings.
( 6) Closely akin to this category are the many in­
,,
'1 stances of inexact dating of historical documents be­
!
cause of the conventions and formalities involved.
For example, the official text of the Declaration of
Independence is dated "In Congress, July 4, 1 776."
To the unwary reader it would appear that those who
signed it were present and did so on that day. In fact,
the formal signing took place on August 2, 1776, and
. some members did not sign until a still later date.15
Some medieval rulers used to date documents as of
/}· certain towns though they were not at those towns
on the dates indicated. The modern official's and
businessman's habit of sending letters on office sta­
tionery regardless of where they may be or of dictating
but not reading their letters, which are signed by a
rubber stamp or a secretary, may make it very diffi­
cult for future biographers to trace their itineraries.
15 Carl Becker, Declaration of Independence (New York, 1922),
pp. 184-5.
 160 UNDERST_<\NDING HISTORY
Ban:k checks, having the city of the hank's location
printed on them, may also prove misleading as to the
signer's whereabouts.
(7) Expectation or anticipation frequently leads a
witness astray Those who count on revolutionaries
to te bloodthirsty and conservatives gentlemanly,
those who expect the young to be irreverent and the
oln crc:.bbed, those who know Germans to be ruth­
less and Englishmen to lack humor generaily find
bloodthirsty revolutionaries and gentlemanly conser­
vatives, irreverent youth and crabbed old-age, ruth­
less G'::rmans and humorless Englishmen. A certain
lack of precision is found in such witnesses because
their eyes and ears a:e closed to fair observation; or
because, seeking, they fi:id; or because in recollection,
they tend to forget or to minimize examples that do
not confirm their prejudices and hypotheses. (This
sort of attitude is only a special kind of bias and might
be regarded merely as a subdivision of Paragraph 2
above.)
* * * for the St. Bartholomew Massacre, Bismarck's satis­
Unwillingness to tell the truth, whether inten­
tional or subconscious, leads to misstatements of fact
more often than omissions of fact. When the same
witness is both unable and unwilling to tell the truth
(as is mostly the case to some degree at least), the his­
torian has before him a document that commits errors
both of omi$sion and commission. Yet he must con­
tinue to bear in mind that even the worst witness
may occasionally tell the truth and that it is the his­
torian's business to extract every iota of relevant truth,
I
if he can.
\.
(
i
THE ·PROBLEM OF CREDIBILITY 161
Conditions Favorable to Credibility
Fortunately there are certain conditions especially
favorable to truthf�lness, and students of evidence
easily recognize them. They are frequently the reverse
of the conditions that create an inability or an un­
willingness to tell the truth.
( 1) When the purport of a statement is a matter
of indifference to the witness, he is likely to be un-
���-
( 2) More dependably, when a statement is preju­
dicial to a witness, his dear ones, or his causes, it is
likely to be truthful. That is why confessions, if not
forcibly extracted and if deposed by those in good
mental health, are considered excellent testimony,
sometimes acceptable in law courts without other
direct. testirnony. 16 111e historian must be careful,
however, to make sure that the statement really is
considered by the witness to be prejudicial to himself.
Cases like that of Charles IX's claim of responsibility
faction with his revision of the Ems dispatch, and ex-
Nazis' or ex-Communists' contrition over their youth­
ful errors come all too readily to mind. In such cases
the deponent may be engaged in a subtle and per­
haps unconscious form of self-pity or even of boast­
ing rather than in confession, and other tests of trust­
worthiness must be sought.
( 3) Often, too, facts are so well-known, so much
matfers of common knowledge, that the witness
would be unlikely to be mistaken or to lie about
16 Wigmore, pp. 305--6.
 162 UNDERSTANDING HISTORY
thern: viz., whether it rained last night, whether a
prominent citizen was assassinated last Tuesday,
whether a famous bishop was a notorious philanderer,
whether a well-known lord had the largest herd of
sh;ep in the county, etc. Whenever the implications
of scch testimony suggest that such matters are com­
mon knowledge - and especially when they are also
commonpiaces - the absence of contradictory e·1i•
dence in other sources may frequently be taken to. he
confamalion. For example, it is a commonplace that
old soldiers are grumblers, and, besides, many persons
had abundant chance to observe this phenomenon
in particular armies; hence we are prepared to be1ieve
the tradition that many of Napoleon's veterans were
grognards even on otherwise iIJadequate testimony. If
that kind of statement had been incorrectly reported,
it would in all probability have ceen challenged by
other contemporaries writing subsequently.
This process of reasoning rests, however, upon a
sort of argumentum ex silentio ("Silen::e gives con­
sent"), and such arguments can easily be abused. Care
must be' taken tc, ascertain whether, though apparently
commonly known or commonplace, the matter under
ex:imination was iu fact so regarded by other con­
temporaries, and whether they ever had a chance to
le:im of and to contradict the earlier testimony. In
times of panic, for instance, it is easy to exaggerate
J-he number of enemies of the state, and the very ex­
istence of the panic may lead to silence on the part
of those who do not share it. vVhere, on the other
hand, there is any reason to believe a matter extraor­
dinary, though commonly known, the argument from
THE PROBLEM OF CREDIBILITY 163
silence would work the other way: the very fact that
a statement of something extraordinary was not cor­
roborated in other sources that might have been ex­
pected to mention it would render it suspect. The
ambivalence of the argumentum ex silentio makes it
a weak test for most purposes. It is not the silence of
other possible witnesses but whether an event was
considered commonplace or extraordinary that lends
credibility to or removes credibi�ity from single state­
ments on matters of common knowledge.
(4) Even when the fact in question may not be
well-known, certain kinds of statements are both in­
cidental and probable to such a degree that error or
falsehood seems unlikely. If an ancient inscription on
a road tells us that a certain proconsul built that road
while Augustus was princeps, it may be doubted with­
out further corroboration that that proconsul really
built the road, but it would be harder to doubt that
the road was built during the principate of Augustus.
If an advertisement informs readers that "A and B
Coffee may be bought at any reliable grocer's at the
unusual price of fifty cents a pound," all the infer­
. ences of the advertisement may well be doubted with-
out corroboration except that there is a brand of
coffee on the market called "A and B Coffee." Al­
though the opinion that "William Jones' widow is
a more charming lady than Mrs. Brown" may have no
validity as testimony regarding the relative merits of
the two ladies, it is probably good evidence on the
physical condition of William Jones.
Even the boldest propaganda may be made to yield
credible information by a careful application of the
 104 UHDERST ANDING HISTORY
rule regarding the incidental and the -probable. Such
a statement in a propaganda leaflet as: "Our aircraft
easily overcame the enemy's," would be, without con­
firmation from more reliable sources, thoroughly sus­
pect with regard to the inferiority of the enemy. Yet
it may be taken at its face value as evidence that the
enemy have airplanes ( especially since it is not only
incidental and probable but also contrary to interest
in that regard). And the statement may also have some
value as evidence that "we" have airplanes ( though
that value is not as great as if the ·evidence were here
also contrary to interest). When in a war or a diplo­
matic controversy one side takes the trou�le to deny
the propaganda of the other, neither the propaganda
nor the denial may be certified thereby, but it be­
sames clear that the propaganda has seemed worthy
of scme attention to the other side.
( 5) When the thought patterns and preconcep­
tions of a witness are known a::1d yet he states some­
thing out of keeping with them - in other words, if
statements are contrary to the witness's expectations
or anticipations, they have a high degree of credibility.
Thus, a statement by a Soviet observer regarding in­
stances of working-clas$ contentment in a capitalist
country or by a capitalist observer regarding instances
of loyalty in a Soviet country would be especially im­
pressive.
It must always be remembered that the skillful liar
can sense these con�itions favorable to credib�lity as
well as most historians. Hence he can manufacture
an air of credibility that may easily take in the unwary
investigator. The existence of conditions favorable to
THE PROBLEM OF CREDIBILITY 165
credibility such as these must first be established and
never taken for granted in any given instance.
Hearsay and Secondary Evidence
The historian, let us repeat, uses primary ( that is,
eyewitness) testimony whenever he can. When he
can find no primary witness, he uses the best second­
ary witness available. Unlike the lawyer, he wishes
to discover as nearly as possible what happened rather
than who was at fault. If he sometimes has to make
judgments, he does not have to pass sentence and
hence he does not have the same hesitation as a judge
to permit evidence that practice has ruled out of
courtrooms.
In cases where he uses secondary witnesses, how­
ever, he does not rely upon them fully. On the con­
trary, he asks: ( 1) On whose primary testimony does
the secondary witness base his statements? ( 2) Did
the secondary witness accurately report the primary
testimony as a whole? ( 3) If not, in what details did
he accurately report the primary testimony? Satisfac­
tory answers to the second and third questions may
provide the historian with the whole or the gist of
the primary testimony upon which the secondary wit­
ness may be his only means of knowledge. In such
cases the secondary source is the historian's "original"
source, in the sense of being the "origin" of his knowl­
edge. In so far as this "original" source is an accurate
report of primary testimony, he tests its credibility as
he would that of the primary testimony itself.'
Thus hearsay evidence would not be discarded by
the historiaIL as it would be bv a law court, merely
 166 UNDERSTAL�DING HISTORY
'I
((
�
beca 1se it is hearsay. It is unacceptable only in so far
1 nesses is obvious. Independence is not, however, al­
as it cannot be established as accurate reporting of
primary testimony. A single example wi11 perhaps
suffice to make that clear. A White House correspond­
mt stating what the president had said at a press
conference would be a primary source of information
on the president's words. The same correspondent
telling a presidential secretary's version of what the
president had said would be a secondary or hearsay
witness, and probably wonld be successfully chal­
lenged in a courtroom; and yet ·if the correspondent
were a skilled and honorable reporter and if the presi­
dential secretary were competent and honest, the cor­
respondent's account might be a thoroughly accurate
statement of what the president in fact had said. Even
the most punclilious historian might retain that kind
of evidence for further corroboration.
Corroboration
A primary particular that has been extracted from
a document by the processes of extemal and internal
criticism so far d�scribed is not yet regarded as alto­
gether established as historical fact. Although there is
a strong presumption that it is trustworthy, the gen­
eral rule of historians ( we shall soon note exceptions,
however) is to accept as historical only those partic­
ulars which rest upon the independent testimony of
two or more reliable witnesses. 17
The importance of the independence of the wit-
ir' Cf. e.g., Bernheim, pp. 195-6 and 544; C. V. Langlois anr:I
C. Scignobos, Introduction to the Study of History, tr. G. G. Berry
(London, 1912), pp. 199-205.
THE PROBLEM OF CREDIBILITY 167
ways easy to determine, as the controversy over the
Synoptic Gospels well illustrates. Where any two wit­
nesses agree, it may be that they do so because they
are testifying independently to an observed fact, but
it is possible that they agree only because one has
copied from the other, or because one has been un­
duly influenced by the other, or because both have
copied from or been unduly. influenced by a third
source. Unless the independence of the observers is
established, agreement may be confirmation of a lie
or of a mistake rather than corroboration of a fact.
It frequently happens, however, especially in the
more remote phases of history, that diligent research
fails to produce two independent documents testify­
ing to the same facts. It is also evident that for many
historical questions - the kind that would especially
interest the student of biography - there often can be
no more than one immediate witness. Of the emo­
tions, ideals, interests, sensations, impressions, private
opinions, attitudes, drives, and motives of an individ­
ual only that individual can give direct testimony, un­
less their outward manifestations are sufficiently well
understood to serve as a reliable index. Even when
those inner experiences are known from the testimony
of others to whom the subject may have told them,
they rest ultimately upon his own powers of introspec­
tion. The biographer is in this regard no better off
than the psychologist - and worse off _if his witness is
dead and beyond interview. And all history is bio­
graphical in part. The biographer does, however, have
one advantage over the psychologist - he knows what
 188 TJNDERSTANDING HISTORY
l
1/
his .;ubject is going to do next. He therefore can
reason from response to sensation, from act to motive,
fro'Jl effect to cause. The compli;ted behavior pattern
n1ay give confirmation to the biographer of the in­
ward psychological processes of his subject.
It follc,ws, then, that for statements known or \i
knowatle only by a single witness, we are obliged to I would believe only that Cellini and his corroborators
break the general rule requiring two independent and
reliable witnesses for corroboration. Hence we must
look for other kinds of corroboration.. A man's pro­
fessed opinions or motives will seem more acceptable
a� his "honest" opinions or his "real" motives if they
are not in keeping with the pattern of behavior that
would be "fashionable" in the society in which he
moved but at the same time are in keeping with what
otherwise is known of his gec1.eral character. 18 The
very silence ( i.e., absence of contradiction) of other
contemporary sources upon a matter appearing to be
common knowledge may sometimes be a confirma­
tion of it (see above, p. 16z). In other cases, a docu­
ment's generC1l credibility may have to serve as cor­
roboration. The reputation of the a.1._1thor for vi;racity,
the lack of self-contradiction within the document,
ihe abse11ce of contradiction by other sources, free­
dom from anachronisms, and the way the author's
testimony fits into the otherwise known facts help to
determine that general credibility.
) dence to a claim of virgin birth recently made in an
Conformity or agreement with other known his­
torical or scientific facts is often the decisive test of
evidence, whether of one or of more witnesses. That
(,
l�·�· - ·
1a Cf. F. H. Knight, "The s;ckness af Liberal Society," Ethics,
LVI (1946), 90-1.
- 1t'''
:�,.
� �;-.
l&a�t;·�-i .,.,..-
THE PROBLEM OF CREDIBILITY 169
Cellini saw fire-dwelling salamanders, devils, h;,los,
and other supernatural phenomena would hardly
seem credible to any modem historian, even if Cellini
were otherwise generally truthful, consistent, and un­
contradicted. And even if Cellini's statements were
confirmed by independent witnesses, the historian
saw things they thought were salamanders, devils, and
halos. General knowledge of ho'?! little effect a thumb
in a hole in a dyke would have upon preserving a dyke
that had begun to crumble would be sufficient to
destroy credence in a well-known legend, even if there
had been any witnesses to that Dutch hero's tale.
Doubt can be thrown upon the old story that the po­
tato was introduced into Ireland by Sir Walter Ra­
leigh and hence to England by merely pointing to the
fact that the Irish potato is of a different variety from
the English potato. 19 What little we know about the
, time sequence of cause and effect induces us to be­
lieve that if notable contributions to anthropology ap­
peared before and around 18 59, the birth of modem
anthropology cannot be said to be the result of the
publication of Darwin's theory of evolution. 20 And,
for obvious reasons, it is difficult to give much cre­
English divorce case.
Because the general credibility of a document can
rarely be greater than the credibility of the separate
19 W. H. McNeill, "The lntrocluction of the Potato into Ire­
land," Journal of Modern History, XXI (1949), 219.
zo Cf. F. ]. Teggart, Theory of History (New Haven, 1925),
pp. l05--6.
-
 !
I ,
:
170 UNDERSTANDING HISTORY I THE PROBLEM OF CREDIBILITY . 171
,�i

details in it, corroboration of the details of a witness's
· testimony by his general credibility is weak corrobora­
tion at best. Likewise the argumentum ex silentio and
'i\ conformity or agrnement with other known facts may
\ pened last year, the sources are many and not always
;.·: 1
known, and the contradictions among them not yet
.,..-.. ,
familiar or reconciled. It is easier, among the enor­
!:'l mous collections of little exploited or totally untapped

i:
11·1
) materials on happenings of recent periods to find,.
·a
be misleading. They are in the nature of circum­ II •
t something unknown to describe or to reinterpret
stantial evidence, the weakness of which any reader "
of court proceedings and detective stories knows. familiar story on the basis of hitherto unused docu­
While, in the cases under discussion, these tests are ments than to do either for events of remote periods.
Hence, as a general rule, the more recent the period

,:·
proposed only for ccmfirmation of the direct testi.­
mony of one witness and ·not as exclusive sources of
evidence, their circumstantial or presumptive nature
·:
l.,
renders them S'.lspect even for that purpose. Hence
histo,ians usually insist that particulars which rest on
a single witness's testimony should be so designated.
Thf:y should be labeled by such tags as: "Timcydide$:
says," "Plutarch is our authority for the statement
that," "according to Suidas," "in the words of Eras­
mus," "if Boswell i� to be believed," etc.
-,"' if
\t
of study, the more difficult it becomes to say some­
thing that will remain long unchallenged; for both
'
the intensity of controversy and the likelihood of a
·.'·'
new approach tend to increase with the proximity in
time to �me's own day. Thus a greater degree of con­
")
sensus and certitude may easily exist among historians
where the testimony is lacking than where it is full.
Perhaps nothing provides more eloquent proof than
this that the historian's "truths" are derived from
analytical evaluations of an object called "sources"

.0., :_ ::
Certitude vs. Certainty 'Ti
rather than of an object called "the actual past."
"'0111 f Since such precautions are not always taken and f-
�.:•
these single-witness statements are not always treated t:1
i:

as probanda capable only of a lower order of proof, a tp i

cnious paradox results. For many early periods of
history, less disagreement is found among the sources,
t)
1· ,
�r

:�it
because there are fewer sources, than for more recent
periods. On what happened one or two thousand
years ago, despite the steady increase in archeological, · Ji/1t
jr
w
epigraphical, papyrological, and paleographical ma­
terials, the sources are few, fairly generally available ·,·;.
and known, and the contradictions among them rela­
tively familiar if not always reconciled. On what hap- ;d '
:
)
>
jt,'i;:- >'·.......
,.r
-w
IL::

; :�r-?."",�

r.. ·