OTHER BOOKS BY ROLAND BARTHES ROLAND BARTHES
A Barthes Reader
oo Ea The Rustle
poe
pe of Language
ee Translated by Richard Howard
(On Racine
The Pleasure ofthe Text
The Responsibility of Forms
Roland Barthes
Sade | Fourier! Leyla
siz
Writing Degree Zero
UW) ILL aso WANG + sew york
A division of Farrar, Straus and GirowsThe Discourse of History
‘The formal description of groups of words superior to the
sentence (which wil for conveniences sake be called course)
{s not of recent date: from Gorgias tothe nineteenth century,
inwas the specific objec ofthe old thetorie. Recent developments
of linguistic science nonetheless give ita new actuality and new
‘means linguistics of discourse may henceforth be posible by
reason ofits effects on literary analysis (whose importance in
teaching is familiar t us) it even constiutes one of the frst
tasks of semiology.
This second linguistics atthe same time that it mus sek out
the universals of discourse (f they exis), in the form of units
and general rues of combination, must obviously decide if
‘sructural analysis permits retaining the old typology of dit-
‘course if itis indeed legitimate sill to oppose poetic discourse
to fictional discourse, fetive narrative to historical. I ison this
Jast point that I should like wo offer some reflections: the
narration of past event, commonly subject in our culture, since
the Greeks, to the sanction of histrial “science,” placed under
the imperious warrant of the "eal," justified by principles of
“rational” exposton—does this narration differ, infact by some
specific feature, by an indubitable pertinence, from imaginary
narration as we find it in the epi, the novel, the drama? And
i€ this feature—or this pertinence—exiss, in what ste of the
discursive system, at what level of the speech-act, must we locate
it In order to answer this question, we shal observe het, in a
free and by no means exhaustive fashion, the discourse of
several great clasical historians, mainly Herodotus, Machiavelli,
Bossuet, and Michele.
1718 Fram History Reali ‘The Disoure of History 139)
Aiferent expressions; they can all be reduced, however, tothe
indication of a movement of the discourse in relation to its
substance, or more precisely throughout this substance, some:
‘hing ik such temporal o locative deities sve ols hence
‘we ave, in relation to the Bow ofthe epeech-ac: immobility (as
tie hav said earn), harking back (lia epetr,rpliare da i
fale ogo), the return (ma rtornando all ordine nar, dice come
ss Ube alt om ds pont, we shal say no mae), the declaration
(er are the ather memorable actions he performed during his rein)
‘The shiferof organization raises a notable problem, which we ean
only mention here: i is generated by coexistence of, to put it
beter, by the conflict of two time spans: the time of the speech
act and the time of the material tated, This conflict gives tse
to important phenomena of discourse; we shall cite three. The
first refers to all the acceleration phenomena of history: an
‘equal number of pages” if such isthe crude measure of time
in the speech-act) cover varying laptes of time (ime of the
‘material stated): in Machiaveli’s Histor of Florence, the same
measure (a chapter) covers several centuries here and some
twenty years there; the closer we come to the historian's own
time, the more powerful the presture ofthe speech-act become,
and the more history slows down: there is no isachrony-—the
result of which i implicitly to auack the linearity of discourse
and to reveal a possible “paragrammatiem” of historical speech,*
The second phenomenon also suggests, in its way, that the
discourse, though materially linea, when confronted with his
torical time apparendy determines to explore this time, pro-
ducing what we might call zigzag history: thus, with each
character who appears in his History, Herodous goes back to
theneweomer’s ancestors, then returns tis point of departure,
inorder to continue a litle further—and to begin all over again
Finally, a third phenomenon of dicourse, and a considerable
1. Speech-act
[And first of al, under what conditions i the classical historian
led-—or authorized—to designate, in hs discourse, the very ac
by which he uters i? In other words, what ae, onthe level of
discourse—and no longer oflanguage—the shifters (inthe sense
Jakobson has given this word) which assure transition from
¥atement to speech-act (or conversely)?
Tt seems that historical discourse involves two regular types
of shifters, The 6st ype we might call shifters of litening. This
Category has been observed, om the level of language, by
JJekobson, under the name timonil and under the Formula
‘GrCHCH besides the event reported (C9), the discourse mentions
both dhe act ofthe informant (C*) and the speech of the “writer”
who refers to it (C*). This shifter therefore designates all
mention of sources, of testimony, all reference to a listening of
the historian, collecting an eiealee of his discourse and speaking
ic Explicit listening isa choice, for itis possible not to refer to
it it relates the historian to the ethnologist who mentions his
informant; we therefore find this shifter of listening abundant
in such historian-ethnologits as Herodotus. ‘The forms they
employ vary from interpolations of the type as Ihave heard, to
‘ou Inowedge, tothe hstrian's present (a tense which attests to
the speaker’ intervention) and to any mention ofthe historians
personal experience this s Michele's cae, who “listens” tothe
History of France stating from a subjective ilumination (he
July Revolution of 1830), and accounts for it in his discourse.
"The shifter of listening is obviously not pertinent to historical
discourse: we find it frequently in conversation and in certain
fries of the novel (anecdotes recounted as "heard from”
Certain ive informants who are mentioned.
"The second type of shifter covers all the declared signs by
which the “writer” in this age the historian, organizes his own
Uiscourse, revises i, modifies it in the process of expression; in
short arranges expic references within it Thisis an important
shifter, and the “orgenizer” of discourse can receive many
1 Fotoing J. rineve (Babbin, le mote log ee roman" Cite,
ae Ati all eae Jong ted ow
Ses neh ai i cm pe of130 From History t Reality
ne, atest to the destructive tle of the sf of ergot
in relation to history cone thi oles naguratons
tt historia douse, places where the beginning of the a
teal sated and the exordim of the speethact ae unted+
The dicoure of history knows, in genera, two forme of
inauguration: fst ofall, what me might al the porate
emg for it opech i actly a solemn a of foundation,
the model ofthis s poetic the sng ofthe poets thus, Jinile
Begins his history bya relive appeal Cn the name of Got
‘Almighty e Jehan, Ste defini nae to be writen the ie
Of our ly King Loui’, tnd even the soci Losin Blane
docs no dian the puijng inet 3 diel does the
Srcepion of speech remain—or so cred, let sy: aube
{etl amuch more common un, the Preface, achaaceriic
‘peech ace. prospecve when it announces dicourse to come,
Srretrspetne when i judges tha dcours (sn the grest
Preface wth whih Michelet crowned his tarde Prone once
it wan completely writen and in fact publhed). Our review of
these unt tends o ngs that the entrance ofthe specerat
int historia statements through ifr of rans, ban 96
its gel not so much to pve the hiorian a chance to express
his Subject” ao “complicate” histor’ chronicle time by
Confronting it with another tine that of dicourse oe a ine
Seay ientya poer tin nor the presence ioral
aration, of explicespeechat signs tends to"de chronologic”
the historical thread and o restore, only a reminnence
tra nostalgia comple, prametri, non ner tine whore
deep space real the myth ne of the ancient commogonis
ito linked by esene tothe speech of the poet or the
{Tee cumiu a dcr
Sra apt 4
fee ing py pen. ove eel ma ny
Stel eof en ne nh nein oe it
sono the mon inereing rbiens
mf ihe antenna
The Discoure of History 131
soothsayer in elec, the shifters of organization attest~if only
by certain apparently rational detours—to the historian's pre-