''

Vhat an unreasonable request! Ten films! ... No, Sir, we are still not
equipped to grant aceess to that knd of stuff ... The library would look
as if it has been infected by some kind of tapeworm.
An employee of the Bibliotheque Nationale in Ps
answering GuiUaume ApoUinaire, 1910
How to Contact Moving Image Archives
Once we have identified the object of research, it is time to knock on the doors of
the archives where tu flms that interest us are kept. There is nothing specia about
this stage: you write letters (no phone calls, please) specitying the nature of the pro
Ject, and ask about the conditions under which it is possible to see the copies you
need. At this point, the really important decisions have been made already; that is
to say, !he methodological questions, which the archives are not responsible fr
solving in your place, have been addressed and you are now in a position to inquire
about prcedures such as the conditions of access to the viewing material. In retur
for the services and the access you are :equesting, tJe least you can do is to teat the
collections respectfully and, if possible, to give practical help if the archivist needs
to dear up, any doubts about the identiv of the fs yo u w J . o &Jt iJ
other aspect of your request.
RUU6
Work in a :11 archive involves an agreement
over the mutu<tl rights and d1ties of
the researcher am the archivist.
llot11 must ensure that the act of consulting the Jilm
will cntribute to a better knowledge of the work
and to its material preservation.
HIt1LJ|l DE DETECTIVE
!.b
Putt ing it as dearly as possible, Wcshould acknowledge that the tChV\8Iis 8C1-
body
who \8paid to increae, protect, improve and cKuCv11the property of the
film archive. If the archive's rules provide fr part Clthe c(cC|)qp to be open to
the public, it has a duty to fcilitate the viewing of this materia] and to provide the
most suitable research tools for this purpose. Some atthivc8are poorer than others
and, therefre, not equipped with some of the technical means necessary for the
study of !D: oHe or more viewiHg tables, a projection room, a library
,
a catalogue
open to the pubic, 8 collection of photographs, posters, screenplays anC other
material concering the production and distribution of flms . .
The researcher must Ke.ep in mind the fnancial limits wthin which some flm
archives are forced to struggle. This does not mean H1al the archivist t8entitled to
make life impossible fr the well-intentioned VisitCr. lf the archive is a public inti
tution, the people in charge of it accept the duI\ of perfrming a specifc public
service. If the atChVc is private, and if its mission st<Jtement says that access is one
of the reasons for its existence, it is probable lhal the arChive receives private or pub
lic contributions towards its operations. In tat case, viewing a fllm may cost more,
but that is all the more reason why the paying visitor should be treated w!th respect
8uU CCO
J
Ci8tIOH, not as an intruder. ]f the latter happens, the best answer is to turn
to the public or private organisation to which the fil.m archive belongs. Those who
give financial support to a moving imae archive must know what use |being made
of their money.
This is not to say that the scholar is an absolute and unquestionable judge; quite
the contrary. researcher's responsibilities are even greater, because Bis on the basis
of your behaviour and according to the behaviour of those who preceded tmd will
follow you, that archives will take a UOcnive or an open attitude towards the user.
The policy of inaccessibility or diffidence which characterised some II archives
in the past was due to the cumulative effct of a number of !actors. The lrst was
the difculty or impossibility of ascertaining the legal status of h18,ofen acquired
without the rig!tf|l owners' knowledge (although most owners in the past lost
interest in |h0t h18 afler their comme1ciai exploit;Hion.) In many countries the
archive's concem was, and often still compounded by the fct that there was no
copyright law, or if there was, it was not applied. rhe second fctm 1as the des
perate strugle to save irnage8which were of little or 110 interest to 2nyone, or were
at best simply regarded as the object of a collector's obsession; but it is precisely
thanK8to the thousands of people who cntracted this virus Ihu! hn archives came
into existence. 'fhe third variable is the lack of manners of son1e people who, once
they have gained access to the flms, tend to abuse their own rights. curators had
not reacted to their excesses with a severity which m.ay u! limes have seemed
unpleasant, fihn archives today would have very little lef to show.
\hen making contact with any archive, it is imperative to be nunc>!,lactful and
tenacious. Yu must gain the trust of the archive :Iu!',u !I | nI always
even when inquiring about such elementary rights as:
!
] rgt JQ information a
bo
ut " owned by the 8rcD!\C, in line vtith the
state
catalogue (that is to sa
y, relative cmpleteness and reliability), and
the
(Q j)Cwabout any !00 or legal limitations afcting their avail-
ability;
z the rI
[
h!of access to viewing mate
ria
l;
the
;I access |O supplementa
ry information relating to objects held b the
archive (for example, the da
te
s o
f restoration, whether 5Oucl!Jn_ came fom
another archive or belongs to a spe
cial collection);
+ the right to carry out research, whenever possible, on copies of a quality as dose
as possible to that of the originals.
For thei1: part, visitors should make it dear that they are fully 8w!C of their duties,
and that they intend to respect them. These duties may be summarised under the
following headings.
LlaI!
Requests should he phrased simply and cC8H}the ideal is a list of flms in alpha
betical order by title, with some basic filmographk infrmation: the date, the
production company, the director. Let's make this point dear: if we ask t o sCc'that
American huwith John Gilbert where the prot agonist goes to w8!and loses a leg',
\VC shall at best receive a polite letter skin_ further explanation, but wc should
not be SulprIsed if we receive no answer at l1.!Ulcorrespondent may be a.n aut
ority on silent cinema (and may therefore Dwthat you are referrng to c !!_
i6d(by King Vidor, 1925), but it is totally unreasonable to expect archive staff to
retnember everything we have not taken the time or trouble to fnd out I ! Olr-
selves.
tct|sou
J< a1CiYist |Snot a researcher in the user's service .. Before applying to the archive
you must be as sure as possible of what you want However sophi5tiLated the C8l8
loguing and however much it may allow quick cross-referencing (rare even
in the most organised departments), archive stalf have neither t.he duty, the right
nor IIc Iuc to do work on behalf of the s.cholar.
Therefre, yGLtshould not assume that a request, 0Iinstance, for flms from U
containing point-of-view shots can be considered precise. Let us 888u;c that the
archive's computerised catalogue a!ows you to search by key words (which is highly
in1pmbable), and let us assume that the cataloguers have been ahle to remember
which of the h!1I!UI I906 in the archive's collection might :DCuUcpoint-of-view
shots. The question then remains: what defmition of 'point-of-view' are yOU talk
ing about? There are many, and it is possible that the cata]oguer'' s opinion is
dillrent from the scholar's. Besides, is it right to conduct research on point-of-view
HISTOIRE DE DETECTIVE
shots in e:dy cinema \ithout vieVng examples of flms in which there are no such
shots in narrative patterns which might have justife
d
their
use? This is a problem
fo the researcher, not f the person who makes the nat
er
ia
l available. Nobody in
their right mind would dare ask a librarian to point out all the bo
o
k
s d
e
ploying frst
person narration.
Cre
The objects the archive puts at our disposal must be treated in such a WY that their
integrity is not threatened: restoring a flm is very expensive, and sometimes te
copy which is entrusted to us is difcult to replace (fr example, the print may have
been derived from a negative owned by another archive).
Notie
Requests for flms and related materials must take account of the workload for the
archive where you intend to work. Pullng an archive print from the dimatised
vattts is abso.lutely not like taking a book from the shelf as each reel must go
through a complex and time-consuring process of retrieval, storage in a cool area
and inspection befre being put in your hands. Any plan f large-scale viewing
(anything over twenty or thirty titles) should be presented with the maximum
notice, thus allowing technicians to find and verif the physical condition of the
prints befre making them available.
Moderation
An \enage' day's work al the viewing table allows you to consult, at a reasonable
pace, two full-length fi lms of fve or six reels, or about ten s. horts of a reel each. This
is only a rough guide, because it is certainly possible to spend a week on a five"
minute trick flm, and to exhaust om interest in a single sequence of a two-hour
fction flm in less than ffteen minutes. \len organising a work plan, you need to
take into account the time that may be necessary for repeated viewing; for setting
up the ree; for putting it back in its container; t: stopping the viewi1.1g machine
to take notes or fame enlargements (if taking photographs is allowed by the ru.ies
of the archive); lbr making sure that a research assistant is avail;ble to repair any
possible damage to the flm; and so on. Asking for nl al1) more flms than may use
flly be consulted in the allotted time involves a useless and expensive workload for
the archive staff and encourages a. hurried and superrlcial viewing experience.
Respec
If the material we wish to see is subject to legal restrictions concerning access, it is
better not to insist that the archivist break any agreement made wi. ti the holders of
the rights or the collectors who have deposited their copies. Understandably,
researchers feel a sense of deep frustration knowing that the flms they want are
near at hand and that they cannot see them. Fortunately, this docs not happen often
!.
5llT l!L^/
you are dealing with silent clncmu. If it does hapen, we must not take it out
Oil the archivist, who can do nothing about it (we are assuming here that the
archivist is ln_oGd faith, and not using this as a pretext fr denying u access lo the
material).
8Cc
Handling original nitrate copies (and all flm prints in general) requires a certain
amount of experience and manual dexterity. In general,. nitrate flms are not avail
able to the researcher fr the O\O\ reason that copies must be protected against
damage Inc\I!a! entailed by repeated access. Only after several visits, and afer
demonstrating with convincing arguments your absolute need to consult the orig
inal, will some archives (and only some) be responsive to your request. If all goes
weH, we will be given a viewing table with manual controls where the image is not
projected, but seen against a source of 'cold' li_ht. Jn this context, it cannot. be
stressed enough that examining a nitrate copy is a privilege which is earned by
showing a fla.wless rig our in the implementation of research methods and practices.
ccdback
The results of research should be made available to t!te archive for intemal use b
the staff. This applIes particularly to filmographic data, which cataloguers are
wuys eager to have, and to information about the condition of the copy. If a print
seems in extemely poor shape, of if we know that the original has wonderfl
colours while the viewing copy is a tded black and white, we might tactfully point
this out, ask when the print was struck (t is nobody's fault if this happened ffty
yc0rsbefore), and enquire if anything can be done to have a better one in fture. If
the copy is excellent, and you tell the archivist, she will be encouraged to keep up
with these high standards. An archivist knows very well that a properly printed
viewing ccmcnLcOn\t:b\\csto the reputation C\thc archi ve vh\ch produced :\.
The Tols of the Tade
This section deals with the equipment used to see flms, and both the active (at the
viewing tahle) and passive (in the screening room) use of archival prints. in either
cas,e we ;ight want to ask to handle - or we might be given the opportunity to
examine- the flm print itself directly, fr its images as well as for any other kind
of in!t+1!1Onit may contain .. This is particularly rei evant for copies printed du
ing the silent era, but it also holds true for any duplicate derived fom them.
Therefre, we shall not consider here ftlr lthich are reproduced on media other
than photographic motion picture flm stock. A rationale for this decision will be
provided later inthis chapter anuat the end of the book. If you do consult films \
other media-out of necessity or choice- you can skip the foilowing pages and go
straight lo the last section of this chapter .. Before doing so, however, it might be
worth at least having a look at the arguments set out on pages !to
HISTO
IRE DE DtTECTIVE

So
here we are: we have the
in
fon
t
0\u>and W0are aDOuLto eKarine it. |
we a
re ai]owed to touch and han
dle t
he print itelf; we have to keep in mind a nm
ber
0\things and have at hand s
o
m
e us. e
fu] it0ms WhIanarchive worthy of the
nmne should have:

Gauze gloves. If the copy ha>

been pr
inted recently, you should W0di them to
avoid fngerprints and dirt on U 0f
l
m stock. If the copy dates Upthe sil. ent era,
!hc gloves will primarily protect your
f
ngers fom scratches and dirt resulting
U0] the hS condition (br
ok
e
n [\0idtI0JS dust, chemical 8uD8IaDccS Cma-
nating fom thc nitrate). Iflhe
gloves are not next to the manual rewinder or lh0
viewing table, always l'em
e
mber to ask for a pair. Technicians usually have whole
bo>es of them ava.ilable.
l'nagifyi11g glass.
micrometer gauge. This is used for measuring the fame's aspect ratio, the siz. e
of the perforations, the dimensions of the frame line and, occasionally, the thick
ness of the celluloid .
.
. ^ cmnparative table of lengths and proection speeds (Appendix ), to be used tor
cakulating nning times.
One or more tables ofedge codes (Appendices J,4and 5) lr dating copies.
Some archives also have a microscope with a cold lighl source (to <lVOid dunaging
the print). Such a microscope allows you to discover details which are otherwise
invisible on the screen and vvh. ich at times increase our knowledge of the 111m's
identity and state of preservation. may seem an exaggeration to insist on the
fo r sc!an instrument, but you will soon realise by using i. t that a small semitnms
parent su. rface like the fame is actually a vast visual universe. Some equipment has
a variable direction light source which allows you to distingu.ish the trade marks
embossed on the flm by producers and distributors (see Plate 18), the traces left by
printing machines, and other characteristics which are difficult to recognise with
standard examination practices.
The Viewing Experience
The fate of a silent film and its frtune with today's audiences are decided at the
mom. ent the flm passes through the mechanisms of an optical projector. -he pro
jector in question is most likely to be set up |rshcnving recent sound film.s. 'fhe gate
behind the lens is different from the full aperture gate (the one used for screenings
of silent flms in a museurn context) used in the silent era, in which the aspect ratio
is generally about l.!`to !.`i (see pp. 4and 1). You wusoon notice whether the
wrong type of gate is in use because, for example, the heads and fet of the charac
ters may not appear in | on the screen, or portions of intertitles are cropped.
In many early tllms the perfo rm
ers are so far from the camera that the
of the
130 SILENT CINEMA
f
a
m
e
s
ee
m to be intact; but with some experience and a sense of proportion it is
fairly easy to work out whether the aspect ratio is mrrect .. In any case,, an intertide
fille
d with
text will reveal any possible mistake: you will not be able to read it in its
entiret
y
,
and the someone wi ll have to remind the projectionist that it only takes a
e
w mo
ments to change the gate. (We assume here that te flm was properly printed
d
u
ring the preservation process .. If it wasn't, there's nothing you or the projectionist
can do about it. This point has already been discussed in Chapter 3, p .. 60.)
h is slightly more difcult, but certainly not impossible (at least from a techni
cal st
andp
oint) to ensure that the projection speed ]s suited to the movements of
the characters. For a long time, 16 fames per second was the most common speed.
As we have pointed out on p. 9, the adoption of the standard speed of 24 fames
per second is the result of a long evolution which ended just at the twilight of silent
film. Before then, speeds could be much higher (for some experiments with colour
cinematography) or as low as 14 or even ten frames per second (fr the amateur
Oko flm, 1913), and sometimes they could even vary within the same fm.
I may be too much to ask the projectionist to change the speed during a screen
ing (even if some archives and festivals do this 011 special occasions). His, however,
quite correct to ask that a silent flm be shown at a speed other than 24 frames per
second when this is necessary. Many modern projectors have built-in speed con
trols, in which case it is only necessary to establish wh21 t the appropriate projection
speed is and adjust the projector accordingly. But even if the projector is not one of
these sophisticated models, a variable speed control attached to the motor wid1
diode connectors will fulfil its task quite adequately. It is neither cumbersome nor
expensive, and can be plugged in and removed quite easil
y
. If projection takes place
in a commercial auditorium, the owner should not be allowed to get away vith say
ing that adjustments to the projector require too much time and money.
What cannot he installed at a moment's notice (but is just as necessary) is the
three-blade shutter, a device which remedies the so-called 'fickering' of the image
projected at a speed lower than 22 frames per second. Anyone who understands a
projector and has a little practical know-how wil be able to make the modification
without too much trouble: the image will now seem to beneft from a more con
tinuous tlow of light, instead of appearing intermittent and a source of eyestrain,
as would be the case with a standard two--blade shutter installed on most optical
projectors.
Then there is music, which was almost always an integral part of the cinema
experience in the silent period, and which a number of festivals and some film
archives entrust to orchestras, chamber groups, organ players and pianists. At this
point the needs of the scholar ami those of the general audience part company. If
silent cinema is the object of a study with ambitions of objectivity, it is important
to know what kind of music was performed al the time .. Extant scores from the
period are usually for piano, sometimes for instrumental groups and occasionally
fr full orchestras (see p. 7). Easier to find are the cue sheets instructing the pia11ist
HISTOIRE DE DETECTIVE
131
or the conductor as to which cl assical or popular pieces had to be performed dur
ing
the show.
These were actualy more suggestions than instructions. Simple improvisation
and
variations on tunes in the public domain were the mainstay of musica.l accom
paniment. Some theatre owners even gave up having musicians in the ;mditodum
and
replaced them with som.etimes exceptional.ly complex equipment fi. 1r the mech
anical
reproduction of sound: gigantic carillons \vith a much vare acoustic range,
pianolas with multiple rol ls, string and wind imtmments driven by a pneumatic
apparatus. Few of these precious objects are pr, eserved, and only in extremely rare
cses can you see them at work in a projection mom.;, a specimen of extraordinary
beaut is held by the Nederlands Fiimmuseum in Amsterdam.
Some archives hav, e restored silent flms by r,ecording a soundtrack onto them
(ranging fom a specially composed score to the reproduction of phonograph
records made at the time of the fm's original release, or an arrangement of the
original musk). The aim of this practice was to encourage the commercial distri
bution of flms which otherwise would have been seen by only a very fw. 'fhe
system has some advantages, but it falsifes the nature of a perormance based on a
dear distinction between an equipment producing images and a sound source in
front or behind the screen.
This practice lies at the heart of a debate bet1reen 'purists', fr whom it makes no
sense to show flms in conditions thai do not respect the original spirit of the
medium, and 'reformers', who are ready to accept compromises so long <tS it is ro
the beneft of the .flm. Everyone, however, is agreed on one rule. \Ilith very few
exceptions, justifed by specific historical circumstances, mus. ical accompanirnent
should be intended as a complement to the flm, not vice versa. Peter Konlechner
and Peter Kubelka, co-directors of the Osterreichisches Filmmuseum in Vienna,
have repeatedly argued against any musical accompaniment at all, in the rwme of
the quintessentia.! purity of the silent image. Their position goes consciously against
historical evidence and is an easy target fr charges of critical idenlism. (ncklen
tally, they are also wary of colour, remaining faithful to the abstract stylisation of
black and white and deliberately ignoring that a silent film is often in black and
white just because that is the way a silen! film in colour happens to have been
served in the archive .. ) 'fheir protest is far from being na'ive,. though. It is the
expression of a legitimate desire nol. to overload the 111m with embellishments
under the pretence of making it attractive to the public, as well as a reminder that
these additions and alterations betray the nature and a1n1.osphere of the or iginal
film experience.
The fact remains that an honest musical accompaniment - neither 'wallpaper'
music nor, at the other extreme, a vinuoso recital- is without doubt more htithful
to the flm than an auditorium plunged in aseptic silence. Especially when a silent
screening lasts several hours, nobody will deny that rnusic puts the audience at tbeir
ease, whatever the reasons that prompted them to aUend. And the student who has
12 5L!!1 L!`Lw^
Pf -h
an
ce O)hearing a pianist who disaeetly underlines some dramatic passage or
st
r
esses
an unusual gag may rest assured that her role as CICUlinterpreter or casuaJ
witness is not in danger.
ln the
rchvc8 BU study centre, whether we like it or not, the interpreter pre
vails ov
e
r the witness; There is no DiIc, no sense of community with the people
in the next seats, and usually 10 big screen: the HU is an ancient paintng,, or
a statue protected fom the weather, whose viewing is permitted only under specitc
( <md defnitely artifcial) conditions. Respecting these conditions guarantees the
8uYY|0the object and facilitates the release of some of its secrets.
Most ar
chives have equipment fr the individual viewing of .films: viewing tables
h
sa
fety copies, hand-cranked viewing tables and rewinders for nitmte and, in gen
eral, for ail flms requiring special care. The most corn1m1 viewing tables come
UOD Germany (Steenbeck and KEM-Elektronik Mechanik GmbH)., the United
States (RGI), Italy (Prevost) and France (CTM). When an employee in the Library
u!Congress (Washington, DC) escorts the researcher lo the viewing room, her frst
question is: 'Do you know how t o use the Steenbeck?' But there is no great mystery
about its use. Structurally, the Steenbeck viewer resembles any oth.er flatbed view
ing table. The instructions we are about to give therefre apply to the Steen beck as
well as to every other machine (whether ve1:tical or horizontal) which operates on
similar principles.
First,. however, it would be useful to recall a couple 0lgeneral principles on how
films should be treated. Not all archives allow researchers to handle JuD reels.:
th.is is a pity, since physical contact with films is a signifcant element of research . .
Where there are no such restrictions, and whatever the rmat with which you are
working, there are nvo things you should never do: take hold of a reel lightly, by the
edges only, as if it were a phonographic record; and unwind a film by hand .. The
lirst mistake w:cost you dear. The middle of the reel will slip down to frm a fn
auhl ;rn\:li' 1k l1\le'lt' t'in, 1'1\' se w1H' U1U\c m \ii1|l\sa:cammr
vuu0laO0 to dust and scratches.
]n the best: organised archives an assistant wit! show us the films, how to handle
the containers (the metal ones are sometimes a little hard to open), how to hold the
reel and place it on the viewing table or rewinder. If the assistant is in charge of
putting the lilm on the equipment, the responsibility is hers alone: we shaH only
have the bother of waiting while reels are changed. lf lVe are alone, 01 the other
hand, first mtke sure that the 111m can actua. ily be seen and has not been damaged
since us!it w:s checked. If you are consulting a nitrate copy, and if the print shows
one of the stages of decay described on pages l/-|^,do not try to unwind the flm.
Stop everything, return the reel to its container (or better, don't even touch it) and
call the assistant.
As for safety copies, the assistant will show us how to thread the film through the
mechanisms of the viewing table .. It is very simple-especially since we almost never
have to put the film through the optic cell or the magnetic head that reads the sound
HI STOIRE DE DETECTIVE
!
we shall soon leam to cope by ourselves. The things we have to knmv and
which the assistat might not necessarily tdl us can
be s
um
med up as fllows.
bQccU
The viewing table has a control which <ldva.nces the at the standard speed of
Z4 frames per second (in some models the running sp
e
e
d can be modifed). The
same contol usually alows the flm to be wound or
re
wound at higher speeds, (
it is better not to use it for such a purpose, especiall y Y\DI copies .. The fster
the f]m goes through d;e gears,. the greater the dama_e it sul:rs .. \Vin.ding and
rewinding a flm is not like using a remote control to rewind a magnetic tape or to
move the digital reader on a video disc.
Some researchers suppose that, in order to 'see' several films quickly, it is enough
to run them at gJeat speed to obtain the information needed at that particular time.
This procedure is of doubtful value as a research practice, even when our object of
research may perhaps be a single shot fom each of the chosen flms. A flm cannot
be 'scanned' like a book;. this applies particularly to a silent flm, the anas1s of
which requires greater attention.
CD9uu
If you must go backwards and forwards while viewing the reel of flm, activate the
control gently. Avoid sudden changes of direction: the H sufrs and too much
tension in one direction or the other can break it.
Utc
|I any reason the film should break, or if we fnd it broken, stop the viewer and
call an assistant Jf no assistant is available, you could splice the two ends of the mm
with joining tae (there should always be sorne at hand); otherwise, you can simply
put the two ends of Lhc flm together, winding the reel on in It8 dit'ction of |^0I
and insert a piece of
[
aper 8Othat the technician ca.n lind the break and put .it right
KCWDUD_
\hen you have finished examining a reel, do not rewind it: the t echnician whose
job it is to verify the stale of the copy after viewing will deal with Ihul. H mJy hap
pen that the reel has not been rewound after an earlier viewing: the iragt WI
then appear upside down. Do not try lo rewind the m: |ht technician do
this. Meanwhile, you can Iook a! the next l1hn on your tis!, or sort out your notes.
HCChu!UCt8
`Hu1 and I'tn) flms are wound around plastic cores o reelholders of various
sizes. It is good practice to make sure !ha!there is a core or reelholder of the sanw
size on the side GI!hctable where the l1m cO0tI:(the take-up spool): this wil >I-
plif the job of inspecting the r|nt and avoid the problem llc hU nOIfitting
l.
5ILL!1 'I!L
U8CKIUI!8CCU!8UC: bCC8u8c !C cCC !8 !CCI_Ct!DC ICC!hC!UC I8 8I_C !8L I!C
uX.
Ad051V0JQc
YhCn8I8rLtn_!CW8Ic!8H1,UCVCu8t8UC8IVCI8C!C8IICK!hCC8UC:(!!C!C8U,
8CUC!IDC88 8!II O! _IcCU hF) C!IC h!U!C IC CCC' !!8II8K8C8U8ID_U8n_CI-
C08ICUSIOUn!!C h mW!CUI!l8ICwCuUU.ICICGVCI,!hC _!uCCUIhC!8C C8UC
!I8U8IICU!CI!C II8C!,U8u8_IU_ I!8UU _CI!!U_ UII! Cn !!C _C8:8C!!hCVICW-
IU_!8b!0\!tCu_!w!C!0 UwI_CUCX!!!UC.LU!hCC!Ch8UU,WhCU
vCWIU_
I8 CVt:, IaS!CU !C CUUC! I!C Hm (I!C!8I , 8CUC!IDC8 8 8!:I C! CU C8UC:)
!C!hC
CU!8IUC C! l!CICC I8 WI!!CVCU!!hC CCI8 C! !hC LU !lCD !CC8CUIU_ 8UU w
C88C I!C)C CC!CCKIn_. !U _CUCI8, CUCUCI Ih8!8 ICC 8!CuUUC!C IcWCuUU
!CC !I_!y, Ou! HCl!CI 8CuU !! C wCuUU CuUU !HC CCIC ICC CC8C]. CXC088!vC
!I_h\UC88C8U8C88CI8!CC8CU!C88C8UU!!CCUU!8!Cn,wuC!CCUuC!CC8CUC88
m8K08I! C88C:IIUu8!ICCuC!I8!C8RU !CI!C IU!C CUI u.
Y!CUyCU!8vC8mC8! :C8.CU !!CCUCC! !CIcC[!h8!I8!C8dy,w!CD)CuhvC
C8c!CUI\8:C!CCIIVCC8UC:),8!C!CvICWC:.C8I!C!hC:CV1Cu8tulC,U3nyICC8
8IC8!IH8II8c!CUIC I!CCCICy8U!c8IVCI8C.!!I!8I88C,:CUCvC!hC UU Cm!!C
VICWC |I8Cl C!O:C!C!8CU CUU:C8chC8 I!C UC!8nI8U8.
LJutOD
!, w!IC v1CW!U_, jCu: CXCIICUcC C8U8 yCu !C IC8ISC I!8! UI8 8!ICu!8 cCy I8
8CCIuUIqUC,CIl! yCu:CC_U!8C !!8! I! l88 U!!8IC HUC:888!cHU IUuI_CU!
UCCUC!C8!CI8!ICH,UCUC!COC!I8!.O!CQ!!V|CWC:IDUCUI8!Cy8UUC8 !8U888I8-
!8n! \UI!8!w8y,yCUw!!!88C_8IU!!CC8!CCU8UU!Iu8I C! !C8CIC8CU8IC !OI!hC
8IC!IVC.
Safet
IU8!y,8!!hCu_!I! m8y 8CCD uUUCCC8S8ty !C u0n!ICDI!, UC 8DCKIU_, nC UIIUK\U_
nU UC |CU 8IC 0:U88IC I1C8 !C v!CWIU_ !8C. !vCyUUC !HC8 8 cC0C 8HCI
l!ICC8ClU!CuI8 C! wCIK,u!C:CuI CwU _CCU88WC88 IhC _CCU C!!C ID
III8D0!!cI!C8I u 8ICI8u1C,IU8ICCm8C! 88lUC!!I8 UIC8C (ICC!ICIC:_8U-
I8cU 8It!IVC8 !w8yS !8VC CUC UC! !8I HOD !!C VICWIL_ ICCu). HIhC H!D I8 CU 8
UI!8!C88C,8mCKIU_In I!8 V\C!UIIy wCuUUC CX!ICUC!y U8U_CICu8, !O: yCu,I0I!hC
lln 8C! 8nUlUUCCU!O!C wH00uI!UIU_.
Never Tst a Fim That Says 'Tst le'
J!C I8! C! jI8c!Ic8l 8UvIcC I8 UCw cCU0!C 8UU !! I8 !IuC !C 8!8I! CXC:CI8In_ !hC
8IU, 8lw8y8KCCIU_IUmIUU!COLC88UUI!!a!8whlc!m8y lUluCUcC!CuRUCI-
8!8UUU_C!!!C!|mWC8\C8Cu!!CSCC8nU8vCIU!8C!:8 !!88UU!!UUCUI88.LnC
cCuU88K. v . !yI8:!I!8Im!8U8I!C: C!\IuS!\U_I!C !l!UI!8C! !8UII!CUCU_!!C
!8vC 8IIU__CU !C CI8IU 8cC8S !C !C C!C8, C8IUCU CW !C C:_8LI8C 8 vICwlU_
HI STOI RE DE DETECTI VE
! `O
scss)OnandlOhYcIacnCVvryprccauII
t
o [
c|IhcbcsIO\!OIah!w)Op
I

On
~
|rbu|nIO tIs dcsIr0cI:On:Yha|s lc 0
u

I
r nUw:
Jhc mal!cr s UaI\chaVc lO clO0s

c

anlac Ihc
fitm hrwJa! ({ sOr
|}
r
whaIItssaidIOLc,andltyIOmaKc
>pc
'
O OuIscns
\bulcs.lhs sOr rgh
aHcr aI. But sncc wc hac rcachcd lhs pOnI, why nOI dO I1lngs prOpcrIy' Yc
sDOu!dnO| s!andbclOrc lhch as ilW
cWrr abOUlIObcgnan au|Osy !OVcr
pOdcsI|sacsIhcIcVa!uc,hOwCVcm
as
a

rcdDy Iadmanpp
!a|cqby I]caIrc
Owncrs, dsIrtb0IOrs and zcaJOus ar0hiY
sIs, a pr
l in a uOV!ng
Imagc archiv s
nOtc tban cVcr a ivng Ob]c6I whtch
ass IO bc
`amI0d sympaIhcti,ay and
rcsQccIIu1! as wc!! aswlh Ob)ccl1Vly
!lhcgrt

Ibcng
vicwcd Iurns Ou|IObca
dscovcrj, OutcHt!sw!cam|lyrcW
ar
0
cd

0Ib
causmO\1ngmagcsatccnIi-
IIcswbichdIdnO!crys|a!!scmaIcta!!)
aI!hc
\
mcwhcn IDcywcrcprOduccd,|!ms
!rOm | hcsI!cnl pcrOdhaVca hIsIO1y, adhsIOrys ancndcssminefeld.
hcDrsI!rapsavcrjIrcacDcrOusOnc.hOwcvcrmuchahmmay|OOcom!clc,
8CDcparIsOI!maynOlbc!Ong!Olhc`OrglnawO:K.Jhcjmayhavc!ccnnscr1cd
!O:a !a!crds!rIbuIOnO1lhchp,Or cVcn bOrrOwcd fom OIhcIlc:. 1hc tac|cc
O!tccuIln a h !m,>l tair!y wtdcsprcad a! Ihc cndOlLhc silent cra,was iOuInc
durnIhchrs| jcars Ol|hcJU!h ccnIuty s!rbuIOrs,|hcaIrcOwncrsand prO)cc-
|Ons1s uscdIO cuI,sp1cc andrc-cd!h.!mshr mc mOslVaticd rcasOns. bccausca
rcchad bccn!Os|Or had dcIcrOraIcd, and hadIObcIcpaccd,Dccausc|hcdrama|tL
c|cc| Olan cpsOdcwas cOnsdcrcdwcak, andt!wasdcDcd a|rOr\aIcIOn1ake
1 cOn!OtD |O whaI Ihcpub!CcxpccIcd, a ccnsOrshp cuL ma havc madc |hc cOn
Icn!s O!a scgucncc sO Obscurc |huItI DcCamcncccssary IO re-edit |hc tcsl OlIhc
U1,andsO On.
But !hss nO|hincOpparcd|Olhcsurgtca1 Opcra|Ona flnn underwent whcn,
sOmcycarsal|cr IhI camc OuI,sOmcDOdy dccdcdIOuI Hbacn|Ocirculation
(U:ca!|c1alOnssuHtcd by5Ic1an ycs tro!udcnI10H !n_, I7I ^, a|cnUIOriOu
n Ihs rcsgccI). Jl mahaVc bccn !c!I IhaI IhcOrgna
b1I!!c ncWOncssOmcl1cshadnOlh!ng |OdOw!lhIhO>cIhcy !caccd,clhcrtn
sI,!c, cunIcnIor raghIcs.Yhavcalrcadymcn1tuncd 9) !hu|a| !hcbcglnu|ng
Ol !hc JJ0> su1nd was addcd IO manyOlIarry 5tmOn: cOmc0|t!. ! ta, O
rcCOgniscIha|IhcmusicdOcsnOIbcOng|O|hcOri_|italhms,buIwho can say hDw
maDyOIhcamcndmcn!swcrcmadc:
\Ocopies OlIrVn Y.Yt!!ul> Be
l
ow the Surce, nduccd b !angc in IJ/,
arc gt>crvcd thc Lbrarj O!Ongrcs i: WsingtOn, I in Onc OI |Dtm,
tcDcD nara|orcadsahcav|yinnIccOmmtn!atuna grImm!|cHDan ropa
ganda >IO1y, vhch s hard O lOIIOw u!c>s
yO; haVcsccn !hc uIh0r n' O1 yOu
nOw lhc Orgnals!u:y.^l thc incrtIl,havc|;cnCH1 hm!hcuuor|>cd
andsO1cOncviewing |hsprnIas |hc|rhrsIencounter with Below the Stuj!ce mghI
m:sIaKcn!y bccvc !haIlhc!D never had ntcr!|!c>uIan
y
kind.
But r!nlswhih havc surVvcd wthOp|
intertitles arc |D Ih|>condlIun ! UD
Or anOIhcr OlvatOus rcasOns. IO1 cxap
]
[
D
;;sc tc coptc> hav bccn
1 36
SI LENT CI NEMA
from original negatives. Instead of in terti des, one cn sometimes see, fr a f
r
action
of a second, two diag,onalJy crossed lines on the frame, or words written i n ink,
embossed or 'scratched' on to the negative, which are illegible on the screen (see
Plte 39). In may instances, this writing refers w the text of the intertitle
(which
was pdnted separately and i nserted at the marked place), the colour to be used for
ti ntirg and/or toning, and the sequence number of the shot. Sometimes the inter
title appears tor a b:rief moment; one has enough time to realise that it is there, but
notenough to read it This is known as a flash title, and is common in copies des
til1ed for dist
r
ibution outside the flm's country of origin. Production
companies
often sent their flm negatives abroad without i ntertitles, but accmpanied by pre
cise i nstructions on their number, placing and content. The camera negatives kept
by companies in case of any possible further commercial exploitation were often
without interdtles as welL Some years afer the Russian Revolution of 1917, when
the vaults of most large companies came under the direct control of the state (and
hter of Gosflmofond, the national flm archive of the fmmer Soviet Uni on), the
films produced by Aleksandr Khanzhonkov were confscated just as they were, that
is, a negatives without intertitles. Most copies of films by Khanzhonkov which can
be seen today are struck directly fom a frst-generation negative, and i f the posi
tive prints have i ntertldes i t is very likely that they have been reconstructed at a later
date on the basis of scripts or other contemporary material.
Telling the diffrence between original i ntertitles, those remade a! a later date and
those reconstructed in the archives requires a careful eye for graphics and style, and
a thorough knowledge of 'original' prints. Archives sometimes try to reproduce the
i ntertitles by using the same typeface as the original, or printing a survivi ng legible
frame several times until the intertitle reaches the requi red length. The dimrence
i s noticeabl e because the text will not show the slight scratches and printi ng defects
whih characterise a complete in terti tie, hut wil appear instead as a 'frozen' image
(hence the term ' freeze frame')
Given these practices, two challenging issues recur: is the print compl ete, and is
the flm close to its 'origina:!' condition? The frst question can be answered satis
l:act:oily, al though not definitively . . To the second, however, we can only say: "e shall
never know, especially i f the print i n question is unique. The margin of doubt is
very wide with films of the 20s; there is no definite answer for copies printed befre
then, even i f all contemporary documents agree on a flm's length, and a list of shots
and a tinting and toning record is available. W might find three prints of a film
which are apparently identical, but such an occurrence is i n fact so rare that
archi. vists take for granted that every copy of a given t i tl e has its own distinctive
look.
Between the ti me when an employee of a prod1ction company noted down the
order of the shots and the time when the audience saw them on the screen, any
thing could (and ofen did) happen. Directors and producers had second thoughts;
something did not sui t the censor, who demanded changes; di stributors received a
!UnUt8DCvU1|l). th3l,1!v"
:IU1P`D, hD !5 OB11:OL .Vcn 0Oi1 ''`
' !Iut1 !Ic8 :n Ihrtn
u<tt<8 IC!US0 8tRHOltGv
u3_C83DG
<uUI8
8\cD:tuUc
11
.
t!1 ;,(O)yOItDc]I|B'
u: :DlcI WU
!hc3!lc8IOU8 uS ``~'
c\DSIaDt5UUl !O|v-1>,OH)Out8tS
OD;vcI t !h0O[c vO<OUt\cO,
!h3l lt 8
U\\!J) 3|]ca i he a nninllf''" exe1Tie
regard the cinema as an InSI:umcnI serving their ideas, rather than the other way
round, may greet wI1h a dismissive shrug. However, behind the discrepancy
between t he difrent prints of a tm there are multi pl e inl1uences, perta i ning to
di ffrent cultmes and profssions. These shaped the film in the course of t i me, turn
In_it i nto the object befre us. Moreover, llc h|t DIIhuI object does not end with
the process of preservi ng the film or wi th i t s showing in an archive: restorati on,
too, is the expression of a project, a declarat i on of intent, a relationshi p between
the flm and the peopl e who 'saved' i. t or t hough! they u ld. ^ fim . is never com
pletely finished LC restored once and forever:. t here is always ' work in progress'
around it.
bQ1Cc8
All t hose whO haLUl0d prints of si l ent fi.!ms ( for edi ti ng, censorshi p, or projection
purposes) have almost always left traces of their passing, wh|h can be i nterpreted.
An origi nal nitrate copy usually shows several splices, made fr at least three di f-
1 d
;| LL| '| LA
frent reasons: the fl.m broke, and had to be mended; the contents of the flm were
altered; the f.nished product could not be obtained without separating and joining
difrent parts of a print.
The first reason is self-explanatory: every handiing of a print involves the risk of
breaking it. That risk was particularly high at the beginning of the century,
when
technology was stiU lD its infancy, but even today flms continue t o be
scratched, broken, deform.ed by those machines which Vincent Pine,. former
cura
tor of the Cinematheque franyaise,. called - in a rough but effective phrase - 'flm
manglers', that is to say; flm projectors. His also obvious that, if someone does not
like a flm, or i f part of it is the source of aesthetic, political, economic or moral
concers, the simplest way of modifYing its contents is to tamper with te shot con
tinuity by suppressing, interpolating or adding bi ts of flm.
The practice of editing for creative purposes is less obvious. In the period of the
so-called 'early cinema: the idea of splicing together different shots in order to obtain
a sequence with its own meaning was a real revolution which met with
perpie.xity
and resistance before becoming common practice. Until the time when cutting tech
niques beca.me an art in themselves, complete with rules and tools fr the job, the
pioneers of cinema resorted to the most bizarre methods fr making splices. All sorts
of things were used, from sewing thread to metal staples (see Plate 21) . Very soon,
however, people leared that the most practical way of joining two ends of a flm
was to glue together the ends of each secton.
Thus, even when it became easy to get several metres of continuous positive fm
(which did not happen immediately, since the processing equipment at the beginning
of the 20th century had only a limited capacity), splices fulflled two other fnda
mental functions. These were not directly linked to the principle of'editing' in the
current meaning of the term, but were somehow related to it. The frst was typical
of the 'artifdaUy arranged scenes' of Georges Mlies, dominated by sudden sub
S!?'.!.OtOabl: C p c <d b etnrordinarily complex visions o( IaD|3.
In his memoirs, Mtlies Vrites 1J1at he discovered the secret of these 'special effcts'
in Paris, on the Place de I'Opera, when his camera malfunctioned whi l e a carriage
was fmnt of the lens. When the machine started cranking again, a hearse had
replaced the carriage on screen: this 'tnsf(mnation' looked Iike an amazing optical
efct
The anecdote may or may not correspond to the truth, but Mehes cannot have
taken very long to lear that the trick of 'stop camera and substituti on' could not
be produced successfully by simply interrupting the shooting in order to change the
position of performers and objects: at least one over-exposed fame, correspond
i ng: to the pause, would have appeared on the print, to the detriment of the efct:
instead, it was better to develop the film, cut the blank fames and join the two
pieces of negative exactly at the desired point (during the frst decade of the cin
ema's existence, th. e same result was ofen obtained by repeating the same join on
all positive prints) .
HI STOI RE DE DtTECTI VE
! oV
well a being motivated - particarly in non-fction fm - by the necessity
of reducing or concentratg fe duration of a natural event or a human action at
the moment of proj.ection, the presence of splices is ilso assoated with the colour
ing of the fm stock by tinting, toning or mord<mtlng. The iH segments were
divided according to the colour that had to he applied to them: 8 shots to he treated
with red were immersed in a tan containing liquid p with red dye, all shots
meant to be blue were put in another, and so on .. Innovative filmmakers could cre
ate efects of changing colour within the same shot through ingenious systems
which,. however, could not be applied on a n industria! basis; as a rule, a IruH of
women woulU work Iit splicers and fm cement, placing every shot in tl1e right
place accordi ng to precise instructions. ^ curious German short comedy; ]lagodie
einer Ura
uf
hnmg (O.F. Maurer 1 926), also known as V(nn die Fii1nkleberir1
geb1.1mmelt hat . . . (Ven the Film Editor Got Confused .. . . ), describes the distrac
tions of a female cutting worker and the conse
q
uences of her daydreaming
(upside-down i ntertitles, pieces of fJm that have nothi ng to do with the story, dis
rupted time sequences) .
For those not present at that time who still want to kow \v0Itla peculiar edit
ing strategy is due to experimental daring, to a projectionist's catelessness i n ' o
or to the cavalier attitude of a flm curator i n Vb1 , there is nothing D! it but to
examine these joins directly, a11d IU try to interpret them. How? First, by looking at
their shape. For example, if all the joins of a VU nitrate print appear as straight
li1es, except for one where the comers of the flm have been cut (the joint h< then
a trapezoid profle), and if things do not tally preci sely al Ihat point (the edit i ng
seems too audacious t:or the time, or the story becomes confused) , then you can bet
that the problem does not lie with whoever assembled the copy lI V0, but wib
someone else at a more recent date.
It is fairly easy to check the diference on a mod.em 35mm viewing print as well:
the profile of a splice made on the negative always appears as a thin light l i ne, <llong
the upper or lower edge of the frame (see Plate 2B); a spl ice made on the posit ive
cpy of a former 'generation' wiU show up as a dark l i ne (see ll0 23A}. The di f
ference is unmistakahl< even when the pri nt is a third- or fou11h, generation
duplicate - that is to say, i f other negatives have been pri nted in order to reach the
copy on which we are working. Of comse, this is not the solution to al l our prob
lems related to the study of editing practice in early cinema, but it is defitdy
something.
cDl1HCu1oD\u05
Apart from spli ces, a ni trate copy always has something to Itll us if the print is
closely examined.

The space between one frame and the next ( the llame line): t he thickness and
shape of thi s se
g
ment are very VarI0Ulc and are sometimes characteristic of a
1 40
5| LL| L| | Lf^
SQC6ICQlOOUC!JOH CODQ01. ! | Uqh0QQCH |h6I!HCI0DCIHCI8|DC OHJ])8C-
rlCDCH|IO:0 QCl|DI10llUCHIlC0!|OH O IhC H1,

C 8h0Qc G!c 10!C l1SC H0} 08U UCh0l0CICil5|!C U0 Q0I|ICu6I QIOUuC-
1OHtOuQ0H],

C lFC:1QIIOH8 8lC_ IhC CUgCS O U QllH| , UCICCH lhC QcIOI0|IOH8. SODC

ODQ0DlC8 OugD| |hC I Cg0! UuQ1C0IIOH U|hCII U1HS U}QIIH|lHgIhCl H0C8
Oi l1|Il08 |H Ch0I0t|lSand WOIUIH_8ID0|cOuU ChaH_C0CcUIUIHgt !DC !IDC O
QlOUHc|1OH[$0C0IC8J , 52, 61 0HU1QQCHUIX5). hC88!C lCChHIQuC W05 u8CU
Uy U0HuIc1uI0l8 Ul0WHH S|OCK, WhO ICSO:!CU |U cUUCS IUCH|IjIHg IhC yC0I
WhCHIhC QlIH|h0U UCCH H0UC (SCC !0ICS22,. 47, 54 0HUQQCHUX4};

hC U8I 0iC0D0iIHg |hC UlSI0HCC UCICCH IhC QIOHlC UI IhC C0DCI0 |OI HCg0-
llVC) 0QCIIulC0HU IhC8H0QCO|hCQUSl|lVC 0QCIIU1C. Ih18 I8 VI8ICIH |hC QhO|O
OHQ0gC49, C$QCCI0 UH IhC IIghI 8lUC OIhC D0uC,

hC 8h0UOW O|1CQClC0IlOH8 Uu IhC QU8I|lvC Ol HCg0|IVC COQlC8 HOu 0H
C0lIcIgCHCI0||OH|Q8gC49): WhCHSuChUCI0i!5VI8IUCUHIhCCUgc8O| hC[8DC,
HtXI |O IhC QclUI0|IOH8 U |hC COQ UCIHgVlCWCU, UHCn0y UC 0UlCIU C8I0Ui8h
WDC!hCI 0 QIlHI h0S UCCH UCiIVCU U|ICC|}HOl IhC C0UC0HCg0|IVC Ol WhCIDC
|I lS |hC lCSuHOOIC O HUIC |l0H$tl8,
hC HuuUCI,WIU|h 0HU 8h0QC UIhC QCI0IIUH8, 0$ UC8cl|UCU lH \H0Q|CI ,0IC
08O Ch0I0CIClS||Cl HIhC lUS OIhC C0lQCIIOU.
YDUSI 0UU|O |hC8CClCnCHI8 |hC IhICHC8SU|hCQll H|l I5C, 0|hOUghV0I01IOHS
IH IhlS lC8QcCI UCIcCH OH0 L0Huc|ullHg CDuQ0H] 0HU 0HU|hCI 0C HCgllglUC
c|HCI 1 91 5 . I8 WUIIh QUl HIIug OuI |h0lUC|0Il$ lIC |hC Du1C lIHC 0HU IhC IH8cIIQ-
IIOHU0I\VCCH IhCQCiU0l!OH8lU8ICCX0HIHCUUHQIIHI8D0UC0I|hCIlHCOIhCI
lcC0SC. Archival duplicates reproduce - in whole or, unfortunat, only in p:tt -
UCIC1y IhC I|uHC, HOIWh0I l$ uOuHU lI \|hClUCI0I8, IC IhC 0C|u0$h0QC O|hC
Hcg0IIVC8QCIO8IlOHS,0C8OutIlDCSVl8ICOH tOQlCSITUD 001C1gCHC0I1OH.UuI
|l l 8 \ly UlU1vI, l UI |uO88lUlC, IU UCC1QhCI |hCH. lIhC CUQ 0V0Il0UC lS 0
8B00I hIu0I |h0H 35mm l I I8 UC|ICi IU glVC uQ, UCC0u$C HOHC O |hC c0IUIC8
Uc8CllUCU3UVC C0H UC CX0ulHCU ClO8C CHUUgh. LHC 08I ICCOlHCHU0l|OH.
lUI tXuDQC, hc QIC8CH.C OIhCcUgCH0l u8CU \ | hC0Ihc cOHQ0HyIH 1 907 IS
JLI 7
Identification dues inside or outside the frame are
i ndicators t.o be used comparatively, and 0Ic not
conclusive proof of the IIH|8 identity
HOgu0I0HICc Ih0I IhC 1DWC 0CvX0l!HlHgW0$ 8hUI 0IOuHU 1 907. 1VC KHUvIS
lh0||!0 HH W08 HU! QlHICU !H I 'U Ol C0lCI. hCiC l80 gOUU QUU0UIlIIjIh0Ithe
H| 5|L DE DETECTIVE
1 4!
m W88 aCI\81jD0CC\n 1VU/,DUI I01 IC mCmcn!we 8:ClIU 0\\n_ICvCO
IlConU rc0sOn8U!CUC\DI n

rWCr
IC
4
Ic
!J8nIrD8HCnUcrVeCp
I1c InI 8U8CD: 8UCVC 8 for
U\_ OUI OI0rC8SlIliC8, DUInC!IOIInCt\!ng
I|CIGCn! ty OId 1JD U0{ nU0 '
C_:t0II Iht\8JCIOCuc8 cOp])q;ng
8 ]_UICS1S, !C 8D80: IC
nuUU0r CI 1Icrn8IiVc
h)UIC8v. Pu nIcrcnCc
8Q0018I CUC a ItUCtCj0U8|U CIDDCiCC)tcnIIdCK WCD CnI]VclHC0,
C0UIO IDCsme 8C!UlCn.
8h8dVc IC CC:IUnIIj
C
!"

r
1
Cn\n_ ID8OHOuDllCS8OcC88OH88c
88 WCD CCK|n_dI d HIIIa1C rCC
cUyIC [CCI_ct08!n0nCu8t w:I a dIc
wCrKOsCI8tCH I0IC\0:n8 Jccce lhcD0room oor, Caught in a
(aboret, Gent|e:nen Nc a.H0 !0l|o
u. dl r00UtU Dy lCjSIUnC nJ VJ, OI
WCn8Iru__1n_WID8 I:n!UlID
C
\
dI
" l!JU \It r|l-icn Je .5.Jsus-Christ I u
WiC, UU:In_ICIIUl1X1CnsCgUCnCC,|CSuS rt0Uc8 IDCIOCI8 JU_Ol.\ HUm
IDC I VU/ CGIIUH, 8 Hd1CU IC a CrC88 H IC I V 4lCHaC dHU UcS IH HOu1 OI0
Dac)U:C HUD I0 V V0I81On
dll DdUC Uj IC 88mC O!n0n: IDC WU; J 0
8n_C8CQUCuCC |W:!!WU 8C08).JIS I8Cn8!!n_jUI7C S HUI!m0_:n0Ij.8OHc
I:n_VCrj 8|Du8r IU II :8 IU UC ICUnU, n 0 COUu!V0WH_ j!1DI, dI ICNUScUl U
OUC!nP!I HCW`:
1CsC d!C CXlCDC .3SCS,CIcCu8C,uICXIiCDC dSC8 !VCIH lm0IC\Vc8,und
Wc8a n0VcrCXauSJIDcDdl H UuJlIHC. WWSP IU_O OUKlH_U! |!Ouc,
d1 WCnCcUIOUC 8 88KIUCCn8)I un 8nIUClC_yOIca!yBlH8.PJcIVC8 JVC cnIy
C\0L, 8urV V n_ I|H Ic IIDCS WUcn Cur0IOr uSCU IU juI !C_(I?r Uu 8 S u_C
:CC cVCI)!:n_ Iat 8CCHCU O!U 8uUW1IC CWC!C uudDlC IO ioen.i[ uh0r IC
COr1csCnUn_ni!I0ICI:nI8 8U DcCnlO8I[CIdUCCHUClIDC|3ICyGtSI!UjCUSucC
lHCjWCiCC!CVCU IO c iCHlCall]UH8IdDC, auU I CIC10!C uJn_cru18). !uvclUnl
Viewing Notes
^I I S]U uI, d I8IICL8H8 S IU WIIC GOWH IH dtUui!Sc UI UmjCIc ICIn I C
!DCI10!ICn UcI.V0U !U! ICCXamlua|Ou CI IC Oj) I 8 ujO!IauI 0I0 I! c
V1_:un!,unUCCja8CuSC OldlduCC: ICanUuuIOlWUIuVOVCU8UuU_C8|cU
ID CUr uCCUS auC IO I C!C8ulIS WC WS IG ccVC. ouc u lcqUI!CmCuI dj1C> IU
dHy 1uU C!HI.0CCIu8 WO!, 8uU I0 mcIUUS IU C u8cU In I > CDulcXI 8!C uGI
VCi\ CIH!Cn| !!O1 lO8C UUIC! USClj!:0S. | IC uIguc udIu!C CI Ic UD}ccI
WC dtC UC81 u_ W I S SU I8I 8 m0U1Uut O!_dulSdIUH G IC U8I8 IS 0S>CuI|a,
C8jCC8!jWCH I88 HaIIv!Ua8S0I|u_uU1CS8OuI!1 HSlaSIu_Uuy8cWmtu
UIC8, UiCVcud hWSCcUuC8.Y11IlCICIIdUIIGuUlmjIc8SlOI:1 c!I(C8mt:8yaVC
yIcGCU i8uyUCncHIs, I d 8UiauScCICI!lScGa!IG UVC!UUIc 8Ot I U8jly
u_ SUDC Cc]CuI8I ]iU.CUuICS Da8cU Un I3cI iaIC! IaH HcDO!y. jC!ajS
Ccau8c I8u_ uUIC8 S UH8UC!CC i8ICI jcUCSIIdu O! IU>c WU cuIV8IC I0
U877lH_ uIUIIUH IdI dIScSGuIUlICv(r,IU1lGU|u_,Uijci8jimIUl>CtI
d2HCS3.
+CacIV!I\CIODcCI|0!UuC u_ dIC!S U Hm histoy, JCau ^Ii, lS d a!d
1 42
S | LL| L| MLN^
doxically brilliant result of the former atti tude. Helped by a phenomenal memory
anda remarkable talient as a flm t0nn0l>5tur, Mitry used to amaze his listeners by
remembering in detail shots he had all.egedly seen ffy years before and were now
apparently missing from the print that had just been screened. Quite often it ws
no longer possible to check his claims, but their imperiousness and precision -
together with the occasional checks which could stiH be made - had the power to
sway even the greatest doubters .. On this matter,. what we have said about `cHL:3
filmographies holds true: when the frtility of the interpretation dashes with the
data's reliability, interpretation takes over. Putting the blame on Mitry for having
n1isremembered the development of a sequence or the text of an intertitle, at a time
when historiography and militant criticism were the only wa of legit imising H!
as an art, would be like blaming Vasari for not having used infa-red rays to analyse
Michelangelo's second thoughts and corrections in the Sistine Chapel
Today, the means for analysing second thoughts and corrections do exist, and it
would be foolish not to use them. Some scholars, accustomed to following i n
Mitry's fotsteps (though with none of hi s genius), are hardly inclined t o stringent
reasoning and may try to persuade us that taking notes and cataloguing them is a
waste of tinne, maintaining that hundreds of fle cards and long days spent describ
ing in detail scraps of nameless flms are no use for understanding the inner
meaning of a narrative pattern or a technical innovation. The recurring accusation
will be to compare this kind of empirical caution to te barren meticulousness of
the encyclopedist who pays more attention to the names of things than to their
intrinsic value.
The only reply to these critici sms is to prove they are meaningless, by showing
that names and things do not count for what they are but for what they stand fnr.
Some flm historians may wrangle over the spelling of a director's name, and stop,
satisfied, once the precise spelling has been established . . Anyone might be annoyed
at reading 'Seastr0n' in the frst chapters c!a cC\an `tctOr Sjostrom,. V\cVcrcO
in Sweden fom the early l 9Hls and later emigrated to the U11ited States, where
Hollywood imm(diatdy adapted his name to its own commercial requirements and
to the simplifing power of the English language. ! is, however, one thing to ensure
that Georges Melies is written with tl1e accents in the right place, and another to
believe that inquiry into a film is over the moment one discovers the name of the
leadi ng perfnners. Knowing that a Hu by M1es was made with a prototype
derived from Robert Y. Paul 's Animatograph, rather than with the Cine
matographe Lumiere, is useful to us not because we mav like anecdotes, but because
it i,s irnportant to relate the forma.! and technical characteristics of a film to the
equiprent which was used to produce it and which contributed to its commercial
success ..
The viewing notes are neither equivalent t o a dictionary entry nor to an untouch
able formula. They are a working tool subject to constant revision and updating;
their value is never absolute, but it vulast all the longer the more we endeavour to
HI! STOIRE L DETECTIVE
4
uCC\0 InIrm8t1OnCODpaI:OC!l cWO\DCh8 1uVCOQ\CU
!OtOut:0SuICh (ev0n i:!s l1u !hut1tsuCun In b0 1DucDc0d b]
tD0HC088 O! COU0C!Ing Cu\).IO!:DgQrc\cn!8 u8 ]O nCIing qc18:8WCOchCVc
IO O0O!pOI.Cn!Ia1cC7cC or usefu!kt|O@n_CuI )0pOIy. aDc st1Uj|ng
JgDI\Dg UCIhOC8 IO: inl0IIOt 8t0nC81n !\0D U!8 O1 ID0 1910s: Our nCICs W
!Dt!uCCIHIU81OnOn!DCu8COIOWKCj:gh! hu!WgDCIS0ndu bw worC8on
!h0IU\ul!t8FC0s.8 ttIIDJW28InID0hDandIDaI0:uCIng8eems IGDtmOtc
:08ltaIn0CIDun u8Da:\vIO18yW0MnOlKnOwWhCIh0!In 1IutILu!Cc!a!Wu
OC Vauab0 OtnOt, bD!It 8 WOrlD rcpc)bcrti IDaI nO lcchn)Jl characteristic 8
uD 0nC m I80 unC lhul IhCI0 I8 a. t0aIOn8hI CIwC!1 Iic O8I!iCn!n OI1ghI
sOD1C08,IDC uOV010n!8 OIIhcuCItCss\VIh nI|c8tCnC,unCher 0:OImIng 8IyC.
U I8 up to u8 to UcCIC0how WC wish to !eCOrd Cu:nOIC>. On bc catu8, Jn nOIC-
bOOKsO1 On uOUuICt.VhuIt0uymaIl018 I>huIIh08y8ICHOu0bCIOUoW0U
CIhOCICa]. ! 8hOUC uOW jOu IO UK0 corrections 0as], anC Cut] out quick
cross refeICDL0. The content OI IhC nOIC8 CCp0nds,. OICOu!sC, O1 Ou \08cutCh anC
Ih0Ju0a8 unC UCIhOU> W0 InI0nC to pursue, and that is up IO u8. OW0\?t, sCD0
bu8I InlOtuuIIOn is imQOt!8nI lOt evCt
y
one. bcL:Ist!n IhI> |D!!maIOn, unC
IhCOtdCt nWhCh ImghIaC8tOnIh8VI0wIngnotes., I >nCcCssuIyIOstate Onc0
aDC Ot a! u IIdC IO b0tOWCCnOIOnjOn file cutU>,bD\ 0\0tjIm0 On0huQQens
IO nuuCu m O u80 u 0C0 OIn!OH1uIIOn.
The I:8IlIn0O!Out IC00nOte8is reserved IOIIh0 !UsOHca II'dc,WhChI8Oh0n
nOIOnIhcInI,buIWhI0h n8nyC88c8OuIhI8\ mo!0phvreference QODt LD
ILc80OnUIDcVtIIcIh0II0u8Iappears OnID0cOjOi,Inits ab80Dce, DhC8V
KUL b
^Wuy8 keep cear Ihcd|stnCtIon between .evidence
mund n Ihc prinl and. information drawn from
written sOurc08.
ab0W|Icn8Ou1Cc8.Whatever yOu1preferred system OInOluIIOn - asterisks, num"
O0l> Ot cIICt8 yOu must 0n8uIc that cVcIy |nCItcCI t0 O! information >
accOuanIcCbja bbIOgIahIturefermce. Ijou suspect that |hCllCOnthe !|D|
WJS aCC0CuI8 uIctCuIC,IJXc nOt0OIIunCpeIhJps fnd IhCC8l0 IhI8OcCur!ed. LI
cou8e i t 8u gICaIC0aof trouble, but th|8is Ihe OI|y wuy lOuVOCwh8Ih8jcncd
to uaDj >thOJI> O!ccU IO u8c DcompcIe, distorted OI0\cn wiOng umaIIOn
b0cuu8C Ihcjh8C nOchOIcbD1 to IIu>IIhO>0 whOwent before IhcU.
n gcnctu, uIChIVC8 InCuC0 lDc IIll0> gIven to unidentified lms n round or
8guat0brackets. !IIsnot n0cC88atjIOIOOWIhI>>jslcmto II8extremes, buI I ss0n-
8bc IO CI8InguIsh IhO8C m8 WhCh a8 ycl CO nOl ha\0 J III |IOm lhOsc whO8c
CcnIIj8 8us]CcIOtuncertain. [SOME ARGUMENT] is lhc atchvu title O!JjtnI
aIIhCIal!Onul !Bu1CJCCVsOnPIhVc n OnCOn, unC aOulwhchOnylh0
1 4
SLENT CI NEMA
co
nt
e
nt is
krown: t passers-by start to quarrel because of an artice in the news
paper. T
h
e fact that suc a ti Ue is English does not necessarily mean that IC
wa
s p
ro
duced i n the United Kingdom or the United States, but only that i t is pre
serv
ed in
an archive where English is the ofcial language. (BOTSCHr1FT !!
PFIb) is the probable German title appearing on a print of a Danish flm pro
ducd around 1 906. The flm has not yet been defnitely identifed, but a German
trade
p
erio
dical of that year mentions a flm with this title -whose plot corresponds
to t
he content of the archive print - among the new Danish releases of the week
O
n ano
ther part of the fle card wr:te in parentheses:. (Denmark) , so that we may
have a frm starting point ht our research. Somewhere else note the name of the
perodical
, its issue number and page, or the archive that has catalogued the flm
with these brief qetaHs. The same applies to the date, which can be entered i n the
right-hand corer of the file card: in round brackets if you doubt the archive infor
mation, with a question mark or a 'ca: i f the date given i s a matter of conjecture.
Next to the date, for conveni,ence, you could mention whether the fm is a comedy,
a drama or a trick suject, though t his does not commi t us to providing defnitions
which, especially with early cinema, may be misleading. A tit le defined as `I|uHm
u a tum-of-the-century catalogue can also be a comedy, perhaps of erotic nature,
and can therefore belong in three difrent categories: the distinction between gen
res could change from year to year, at least in the advertising and i n the production
company catalogues.
Speaing of ambiguities, let us briefy retum to the problem of dates (see pages
101-2). YJ

ch date is the most valid when we are faced with conflicting docu
mentary evidence? This questi on triggers endless quarrels. Filmographies most
often use that of a film's first public screening. However, :lmographers often use
the term 'release' for the date on which the flm was made available to te exhibitors,
which is not always the same as the date of t11e frst public showi ng. Strictly speak
\1\_,\v\ .\ V\ a fh>was rde?sed. on 1)anuar
y
l9! l does not necessar1\ mean
that spectators would have been able to see i t on that same day Sometimes it is
enough to know the year of first public showing (or release); mon precision means
more trouble. Any oficiaJ report on the release date of a film is a worthwhile pi ece
of infrmation; but this may not be always found in the trade press, in which case
we have to rely on U

earliest report of the flm's public SCeening or (as far as we

can tell) the earliest published review.
Another common practice i s to mention the date on which a flm was passed by
the censoJs. This usuaiiy preceded the release date by a fw days or weeks. However,
certain titles may have been approved by the censors months or even years before
their commercial distribution (in which case the censorship date is useful in deter
rining when a production was completed}. A. flm may also have gone through the
censorship procedure after it had been seen by the public. Thi s was often iliegal, of
course, but legality does not always regulate ev,ents (and more confusion arises) . It
may also be the case that a censorship board had to examine flms released a long
H| S1C1E Dc Dc!FCTl VE
1 45
time befre i t was itself c5l8D1J8DCU, GJI|aI a p)odu;rwaited whiIcbc!rcma
ing up his uUto present the flm to the censors or to the j 1Iof \0public.
' should also watch out Irthe unfortunate possibilit that a Which went
through tboard of censors and WS 1uGuHCcU,cGJQ0lC 1dt datc HU reviews
i n the D?WSQ8QC!S,H8]never have becHreleased at all. I haQpcned every nGvd)U
Iu.
In order not to |Jvictim to chronoogical obsession (m to an obsCssion |:any
flm data), a reasonable question ve could ask ourselves is: what is the use of te
Uate? If we only need it to QtGV1UC a legitimate refereDce point in time, it Will bc
enough t o specifr that we a!e referring to the
}
C8) the flm QasseU dmgh censor
siQ or the year of general release ( usually tIcsaun, except for films examined i n
CCCIDCIand perhapSreleased i1Januay t he tolowing year). If k)mng whether
a film was distributed in January or in December i S really imQortant IG us ( for
example. because a rival oompany may have pl agarised contents) , then it is
advisable to make a further efort in UetaiL Finally, i f we need to know WUWSIC
HISIto use a technical innovation, it might he more useful to e8tablIsh the period
when the fm \v8S made rather than the date of release or fst showing.
As every historian learns sooner or ldIer,CXCCQlIU18dHUanomilies are so numer
ous I8I8)]ttempt to rationalise events is bound to encounter serious dificulties:
luckily, li fe is more complex than fl mographi es. But since studying cinema also
invGVCSGC81u_WI G8lCS HUHH08,there is no choice., T avoid paralysis, always
measure lC CU!! agaiust the result. You certainly must dl| towards perfection in
cvci
)
lhingyou do, but in this case the foreseen outcome should at least be worth
ICeffort
Thank goodness, not ev,ery aspcct of the viewing lOtes raises such ethical prob
lems. And even the scholr most reluctarlt to engage i n em
p
irical analysi s cdnnol
affrd to ignorC the followiug items:

thC placC ( archive, town, private collecti on) WCIC lc flm comes from,,. or the
circumstances (d festival, a conferencC, a seminar) in which you saw i t;
the nature of the medium (safety,. ni trate, magnetic or digital reproduct i ons);

the frmat of the film or the medi um carrying the images;
I? presence or absence of colour, and the colouring technique used, i f any;

the presence ( i f nec
.
essary, the number) or absence of titles and intcrtltles, the
language i n which they are written, and their nature (original rde.ase, foreign dis
t ribution, reissue, archival reconstructi on) ;
the |yc UIequipment used i n the vi ewi ng session;
the date on which the film was seen;
! 4O
SI LENT CINEMA

IDc number of reels, compared to te overall number in the complete copy;

length and screening time,. compared to the projection speed;

the archive location or accession number;

marks,. symbols and letters on the intertitles or elsewhere in the copy ( names of
acors, catalogue numbers, trademarks and producers' logos).
In that order,. these collected UIon the !? card could look like this:
C!HJ5Sbw,nI),vicintabc,?au/0,r. I , 3 ol3, It40 h, I8 (?4fs),. B7591.
This breathtaking sequence of letters and numbers i s not attractive to l ook at, but
it ofers a great deal of information, saving efrt and possible future mistakes. The
flm was seen at the George Eastman House in Rochest.er,. New York, (GEH), on a
35mm safety (S) copy, in black and white without titles (nt) on a viewing table (we
need not specif the type). On 29 August 1970, the day we consulted the copy,
the iHm was incomplete (we saw only the frst and third reel of a UIc0-tcc fm);
it was 1 ,640 feet long (the viewing table uses this unit of measu,rement U that
archive), and ran about 18 minutes when projected at 24 fames per second.
The location code used by the archive (copied from the box containing the flm)
was FAB 7591 .
Why is it so important t o know aH this? Let's go through the list again. The
archive source is an essential document of identity: we must quote it whenever we
mention the flm (if the film was shown during a festival , ask the organisers which
institution lent it); this is someting we owe to those who wish to check a statement
we made about the flm and establish the source on which such a statement was
based. Not quoting it is a negligence one can barely forgive a mank collector. People
who proudly daim to have seen a copy of a rare or unknown film without point
ing out its source are dishonest towards themselves because they contradict their
own desire to l0l their discovery be known and appreciated; the childish pdvilege
of not letting others see the ob.Ject they are talking about i s not even justifed by the
presumed respect towards |Dc private collector WOallowed them to examine the
priaH, unless collectors spedfcaUy declare they do not want publicity.
Those who are reluctant to say where a print was viewed are dishonest t owards
others as wel l , because they do not allow anyone else the chance to confirm or dis
prove the conclusions they have reached. Leaving aside any malicious intent,
stating from which archive a flm comes avoids confsion and possible mistakes.
Remember that having seen a silent flm during a festival, in a flm archive or a
collector's home does not 11ecessarily mean that one has seen the film in question
i n the form which i s known to others. In writing about our sample flm, we shall
say, therefore, that the copy belongs to the George Eastman House coHection and
i s ,640 feet long, that the second reel is missing, and so on.
OuiUmoving shadows which, for a long lime, coll ectors and archives h1mded out
as viewi ng copies. Hpossible, and i f the research requires it, the ideal is to see every
availabl e pri nt: they may have come from different sources and may therefre reveal
signi fcant difrences.
To note the physical nature of what we saw is also of great importance. Unless
you uIc1U to study the optical qual i ty of images printed on di ffrent !|uI stocks
(di acetate, t riacetate, polyester), :| i s enough to di fferentiate the original nitmte
pri nts from reproductions on safety flm. Y have often mentioned that,. ideally,
consulti ng ni trate copies i s the ultiTnate goal of many research projects on silent
ci nema; but because access to a flm on ni t rate stock is so di ffi cult, you have to make
the most of the circumstances which allow you to do so. A nitrate copy has a d is
tinctive sharpness and texture that cannot be reproduced on other carriers despite
a.ll the progress i n laboratory technology. Moreover, the ori ginal al lows you to
analyse aspects of the fi lm which a duplicate distorts or wipes out altogether: colour
1 O
5| LL !| ML|A
( tinting,. toning, dyeing by hand or mechanical means), perforations, frame line,
edge inscriptions.
The importance of making a note on whether te print viewed is in black and
wite (as in our sample entry, where we used the abbreviation 'bw') or colour
is
self-explanatory. Knowing what colour technique, i f any, was used in the
source
elements is no less crucial, and helps avoid hasty judgments: for instance, a ! >Ub
Pathe ]ecrie can tell us very little about i ts charm if the viewing copy does not repro
duce the amazing stencilled dyes of which the French company was so proud.
Nor
should you rule out the possibility of coming across two or more copies of the same
flm with diferent colour schemes and systems (see p. 27).
For the s ame reasons, you should specif which equipment was used to view the
fi.lm: i n a projection room, on a viewing table (as in almost every major film
archive), or on a hand-cranked rewind bench. You may wish to mention a sequence
which included consecutie panning shots to the right and t o the lef. If the flm
was shown on the bi g screen,. perhaps afer several shorts, you should be cautious
in trusting your own memory: even an expert researcher can lose concentration for
a few seconds. your notes make a dear reference to three consecutive panning
shots, and i f you have seen the flm on a viewing table, you have every right to feel
safe. By the same token,. the credibil.ity of your assertions on t inting and toning is
likely to be greater if you say tat you were given access to a nitrate print on a man
ual rewind tab le.
Between the note on the film format and the equipment used fi,r viewing is some
infrmation on the interti tles. In our sample entry we find 'nt', meaning the print
had no tides at all If there are intertitles and they are wri tten in, for example, French
or German, t he letters 'F' or '\' for wiU remind us of this. If the film was produced
in a country other than the one in which the copy was distributed, the translation
of 'the i ntertitles may not have been accurate, or the text may have been modified.
The number and position of the titles may have been ch;mged, perhaps to suit local
taste or censorship requirements.
The date of viewing is, among other things, a useful reminder of our current
competence. Due to inexperience, some entries in om fles may contain only scant
observat:ons and approximate remarks. In August 9U, for example, we may not
yet have learned how to evaluate high-angle shots or rapid alternating cuts. If these
are .mauers whih now concem us, we would do welll to see the til m again (and until
then include it only provisionally among our research s.omces). l f the tile card dates
from last year, we may perhaps be .more confident that we took note of high-angle
shots and allernating cuts, if we fund examples of either .. It may also be that since
we last saw the flm, te nitrate has started to decay and the archive staff had to Ci>
card some footage in1 order to save the rest of the print. If other people notice
discrepancies between our description and theirs, the date we wrote down may help
us to find an explanation for this.
M.ake sure to note the length of the print: in metres or teet and i n reels ( ideally,
H!I,STOi fE DE DETECTIVE
1 49
the note shou!d make a distincton bet1Veen the total footage and the segments
added to the film after its period of cmmerca release, such as archival titles). Say
ing that a flm in two reels is incomplete is not enough: we should sped!)' which
reel is missing, and keep that in mind if some aspect of the plot is unclear from the
notes taken at the time of screening .. Most viewing tables have a footage counter;
some, like those at the Library of Congress, can measure the iilm t o the frame a.nd
calculate its nami ng time at 24 or 1 8 fames per second.
Working i n a projection room is a difrent n1atter. All we can co there is to vie\V
the flm wit a stopwatch in hand. But this is distracting, impractical and, most
importanty; inaccurate, especially if we do not klO'I' what the correct projection
speed is. It is better to apprach the projectionist and ask i f the cans containing te
reels bear any indication of the flm's l ength. Be careful: the label often cari es only
approximate information, useful as a refrence poi nt but not uncontestable data.
And i n an early flm the diference of half a metre can mean a l ot, both for the nar
rative structure and for such technical and stylistic aspects as cutting and camen
movement
The last item in our sample entry is the location number of the pri nt That's easy
if you a:re working on a vie1ng tab1e in the archive's premises; if you are i n a pro
jection room, all you can do is try to meet the projectionist and ask to write down
the l ocation code and number of the reel being shown. It is extra work, but before
considering i t useless and finicky, it is worth listening to the fllowing cautionary
tale. Among the countless treasures of the National Fi l m and Televisi on Archive i n
London there i s a fil m probably directed by PhilJips Smalley (perhaps i n collabor
ation with Lois vber, or perhaps by Lois Weber alone) called Susen.e (91 3).
According to a fri end, i t i s < extraordinary film, and on our frst trip t o London
we manage to see it Our friend W<ls right. Suspense is an amazing display of tech
nical i nnovation and genuine narrat ive tension: shots div:ided into thncl trangular
sectors (with a diferent action taking place in each), bold posi ti oni ng of the cam
era, a breathtaking edi ti ng style. There is, however, a wrong n.ote: about halfway
through the fl m, during a car chase, a man stands in the middle of the mad and
li ghts a cigarette. Surely he i s about to be mn over. But nothi ng happens: in the fol
l owing s hot the chase continues and the passer-by has crossed the road sate and
sound. How could a remarkable fil m l i ke this be spoiled by such a gross oversi ght?
Maybe the fil m i.s incompl ete. W go home, talk about i t with people who have seen
it, and our accounts do not taHy. Somebody gives a di ferent account of the stor
y
:
the passer-by was run over, and the chase went on. We seek assistance frmn the Cat
alogui ng Department and the mystery i s revealed: NFTA has three copies of
Suspense, but onl y one of them i s complete. W saw the most incomplete one, whil e
our fri end saw a version which was not quite complete. The truth is that one of the
cars brakes and is unable t o avoid the crash, yet the passer-by does not get m.ore
than a fw scratches ancl a fri ght. The pol ice hel p hi m to stand up and go to one
side, and the race carries on. Clearly, it is not our fult that the copy we saw was not
1 50 SI LENT CINEMA
thcmost complete, but once we know which one i s, when we next want to see i t or
show it to someone, it will be essential to ask for the archive number correspond
ing to the complete version. Perhaps five years ago the only copy available was t,e
incomplete, and later a better one was found: when writng about that flm, which
copy shaH we refer to? The only solution is, again, to clarif the identity of the copy
in question.
Now that we have dealt with the soune, the carrier and i ts format, the length,
the date 8)Uthe equipment used for viewing, the presence or absence of colour and
intertitles, t he archive number and the screening time, we have j ust one line of notes
in font of us. But a wealth of information is concentrated into that single line, and
many issues have been raised in this simple attempt at putting together the precise
details of the object we have examined. It may have seemed an onerous job, but we
are now in a much stronger position. With these notes at hand, we can launch into
the most daring interpretation knowing that our fet are frmly on the ground. The
notes must now be i ntegrated by our own ideas (whatever they may be), and ideas
directly reflect our critical skills, hut everybody i s now given the opportunity to
check them against the evidence that has been provided.
Ideally, a film should first be seen on a big screen (that is to say, i n the conditions
closest to those for which it was designed), and then on a viewing table. There is an
instructive qualitative difrence here . . In d1e frst instance, the experience is 'syn
thetic', its purpose being an acquisition of the 'global' sense of the work. In the
second it is analytical. At the viewing table, the flm ceases to be a homogeneous
tlow of images and becomes instead something we can fragment, concentrate or
distort every ti me we decide to stop aIa certain shot, examine the reels out of order,
or study a particular sequence for perhaps a whole afteroon. Such distortion can
not be eliminated altogether, but i t can be brought into proportion. If we have no
opportunity to see the film on the bi g screen, we shoud frst watch it on the view
ing table without interruption, and then start again. In doing so, we must always be
awarc !haI!hS D OO!It_ 8I IuS !O UO with Hc an
audience :DU0:l1uH!a!| f818\Oj,!?8! Ot8U 0DUreverse the 'ourse of
!ime h: uroscs,8HU Ih0|' hn,DU\ I6 a luxu: a CU!SC) N!cW-
C15 O the SIIcn| C:OUJU HO! afrd.
1DWCVCI d1\\HCu \\ U8y OC, V!tW:D_ 3 l1D i U1C J of a J
I\nl is. UChIIC[iClti8D lOVu\cHt_ I\D\!tJ K:DUS Dlrtdia, t!)r n8_-
nC! O:Ul|I8Th t]cUUcS HDI 5C!C Cct!tOD\t |ua_C D \DC Wd \\ Stc !HC
QO\O_!8Q\+ !U8_c, ID3I1 UCCOUC8 DCDO:!$c$t0DC undcrsIandsit difr-
cUly bD!! n teI8t!oH t \\i0 8DG S|8 <C 8HU 0 \tUf1 jSyUOO_18l 8CH8. A
nDD-hDIDi8D!0 8I!!CI can be a. V: DC[u.l 0 irjC0!U,
\O:OU_ 0H0:ySt8of tO)C!cU :iu_C, H!:lUOcS uUIeQ|8c :JDH!c !HI
!lScU.
JUSCWO O]]O5 I:S tD0tC8iC kftH\O
|
C:D\D1! I1u! a\c113!\VC I:1^IS
8C [0lu and economicaL Jhcycha!cny- JUc O\DtSj!d\USCuSCof
H 5!I| LL LL!!V|
1 O!
democracy - the socaUed 'purist' attitude of thos:e who |Oscc irna
g
es o n
a QDCID_I0]!1carrier. (Accoring lO \h18opinion, the IC|that digital On1agnetic
reprduction is cheap and available to all more usc)amaterial of bet
ter quality available only to !e fw; Usubject to wear and tear). But tis argument
is more
a\UO:!\a1antan the assump|ion \\!s supposed to challenge, si nce i| |Mcs
(gr
anted that the distorted imitation of a work of aII is more or less as good as
the work DlaI! i t8ell.
For a completely diffent reason - that is, the understandable efrt to reduce
the costs of replacing wom-out copies - some insu|\|iOs tend to support this
scrious preudice, taking as an alanning justifcation the tct that most users are not
as committed to analysing the formal qualities of the n1oving images as their mu
rative, psychological and 80tIUH!u18 implications. In ethical u 1U ii1telleduai
terms, this i s an unacceptable position, which is bound to have a devastating influ
ence on the future development of film studies; i t places the flm archive on the
same level asan imaginary museum whose curators may decide to show the visitor
photographic reproductions of painti ngs., rnaintaining that 111ost of the public is
solely interested in what appears in the paintings themselves, rather than in their
formal va.lues. The fact that archives Vusoon be likely to have no other choice than
to show flms in the frm of non-photographic duplicaes is not a good enough
reason to just i f the passive ac
q
uiescence to a practi ce which is morally question
able, and i ndt:f ensib!e from a philosophical standpoint.
Since we spoke of 'ideas' about flms, it is worth saying a l0wwords on the best
ways of gatheri ng these ideas. Everyone knows about taking notes by the light of n
pen-torch, one of those which do not disturb the person in the next seat. D1lwhcI
UC fil m only lasts a few mi nutes, or even a fw seconds, everythi ng happens Vry
S,and nCte-taking my not be enough. For sdting down momentary i mpressions
i.n the audi torium or uI t he viewing table, a smal l portable tape recorder is of unde
niable value. Instead of the effort - |n the dark, with the rou:n- of findi ng
the ri ght words to describe what you have seen, al l you have to do ls whisper some
words i nto the microphone; later, these i ntui tions can be pmperly developed | n
written form. You should be guided here by practcal ity and good manners: a nmm
bled sentence wil be diffclt to puzzl e out when replayed; i n order: to speak quietly
but dearly, tactful VCwcr8 try to fi nd places a | Utway fom the rest of thc audn
ence, fr from those who want to enjoy the flm i n peace.
If it is used unobtrusively (it is never pleasant t o be sitting next l someone who
is muttcring esoteric remarks into a pl ast| c box, all the more so i f !1cIc are :tvcI
whisperi ng vC|.cs pollll!i ng the silence in tht theatre), a tape recorder can save us
from creative paralysi s: the brain always runs faster 1J1an the pen, and it is sad to
stop a!the second fil m because scribbling in the notebook has reached an impos
sibly lran!icpace. D i mportant objecti1ons to this method ought to be mentioned
here. The frst is a technical l i mitation: i f the film is shown with music in the ;di
torium, you are n serious trouble, and there is nothing lcf fr you to do except
152
SI LENT Ct NEMA
leave the front rows of the auditorium, sit as fr as possible fom the screen, and
speak a louder into your recording machine. If te music is being perfrmed hy
a .large orchestra you may as well give up, and i t i s better not to take out the tape
recorder if the theatre is crowded and your activity is bound to annoy nearby spec
tators.
The other problem has consequences which can last long after the moment of
projection. Listening to your taped observations in the comfort of your own home
is very sensible, but if you do not deal with the tapes as soon as possible, you wil
end up cursing yoursdf for having used this method at all. Written notes, even when
laconic or scrawled, can usually be deciphered. But a tape is a piece of plastic with
a limited l i fespan, l abelled with ambitious yet anodyne t itles such as 'Danish flms
seen in Pordenone ] 986' or 'Brighton 1978/Pathe'. Before the tapes pi l e up - and
they do -you must transcribe them immediately. You wnot regret i t. If you can
not keep pace with the taped notes ( a whole afternoon i s not enough to transcribe
and develop the contents of half an hour of tape accurately), it i s better to forget
the tape recorder and revert to pencil and paper. We may take fewer notes, but we
won't run the ris.k of being unable to use them when needed. Even if they have not
been reworked or copied out into a more legible form, we can keep tem in fold
ers; our only worry will be whether we can read our own handvvriting m011ths or
years Ia ter.
The swif progress in digital technology s uggests other methods and possibilities
of recording and organising your noles. Copying the recordings directly to the com
puter by means of voice recognition programmes will help you to get ri d of all the
hassles described above. Still, you will need to disentangle your improvised prose
and make things clear to yourself through some sort of editing process. Do it right
after your viewing session; i f you don't, ev,en your computer screen w be a dis
heartening mess of textual fragments.
If this issue has been reolved, all that rema. ins to be done i s to decide how to
manage the different files you have set up for comparative analysi s. You may have,
|rinstance, one main fle for each film and a set of subject Iiles (fr example: fJrns
with onscreen depictions of physical violence; panoramic shots of 360 degrees;
films with the German actress Anna Miler-Lincke; prints coloured with a combi
nation of di ferent techniques) on which the tides of the corresponding films have
been entered. According to the scope of the project, a11d to the extent of our interest
in silent cinema, this data archive may include a few dozen or several hundred fles.
frame Emtlargements
Wi th tbe curator's permission, we may want to bring a camera suitable for making
fame enlargements from the films being viewed. Basically, it i s a slightly modified
version of a slide duplicator \vhich can be connected to the lens by rneans of an
adjustable ftting (there are several models available, including some reasonab
ly
priced ones). Learing to use it does not take long. Yu need to know the optimum
H| o1L| |L LL DETECTI VE
1 Oo
distance bDC IhC 8 unUUC \I 8ou!\V (a nUVl IOthe V0V'IL table
pg

be 8u\table ICI DuCK 8nC white hOICta), CIhcIWI8C 8 l88u Un bc

required rOCuJCUQ1C8UCn),8nthe
0O8uIc!mv1 ieatHCn\OI sens\tViIj
C1IDcflm you are u8ing. More so
p
histicated equ\mcnt is au8bclrHC8cmak
Infame enlargements fequently, or by C1I8!RC8n8.!is Cn8Vc,buI10u\iVc1j
C88j tO C0t0IC, 0n0 83U!tICi\ !c8u!I8 C8n bCCl8:utO even Wh01 ded!In Wlh
CCiquality l 6mm flm. The DO8I8CV0uC0OCrtdDe HO00is cm;ty DCUc0C
by anCn.oCUcu:ChVC88rC0gUQCuto make !t8UcCU!a10uCn\$W1homuter
8c8018. \iCC:iDtcU u8yb0 8n 0XCn8V0 |b0c8u>c of s|uHubCu) 8nUHU0
CCn8uUn8DI, 8!CW8cv01amCnI8|idelivery), but IhI8ClO1.unsometimes
bC Ir HO8I practc88O!uIICn ICOuIrb0ms.
LOnIcI upset, ICu,)yCurinQui:y gets a neaIiVe r08On>C. !81Cng lm
!Diu 81t en8igCmCnt m8 endaner Is pscu1 n!C|I, 8nC In>lIu|On8|!>I
Can1 8UD!C the I8 of !1n Ici WC|K iuncO b a careless treuImenI o Iheii
o!d1s. !I:uI8Cvery uncleUiWh0I0:CI nCI the ubcuICnCa!18Uc enlarge
ment Iequires CCjIhI CC8i8nCc. haI8 nOt a rOblem Iht flm I>in the ublic
CCm8n (88 8i HC8I 8I1CnI Hm8 muC0 u1I IZ UnC0r I0 current laws in Ih0
l1I0CoI8IC8),buIit CCinI I0y 88CI0nI1aIone in aU other cases accOuinto 8Cmc
very cautious CuuICI8.
P8 C| the man8gCment CIhOICiaC I1C:, UIOVJ>8I\Cn s OuI CIh0 guc8-
I1Cn. WI ICn rolls CnCuIiV08 WCh 8VC nOI bCCn 8IuICucC, UcmO!y 8 nO|
0nCuh.1saYC our mernory ibCIl01 u80, IryIh0 flCWin.
OIcIC phOtC8OJ1 the viewing file c8iC>,anCgive a number to lht negaIiVc film
unCte pictures taken;

|nce the negutves hu\0 bC0n InI0C, keep them i n numbered envelopes;

Na|c a list of nC8IIV0s, wi th the photo numbcIs und Ihc IUcs O the cO!Ic
8CnC1n flms;
\enever you print 8 jhCIo, write the negative number on the back. ]n that way,
yOu wi!l be able quickly to locate the hCto lrom the hImyou are after and i Ii
be easy to retrieve the title of un image that slipped out uIyCu fle or that you
cnIIC sCI0CnC. I is nC|nece>sa|y to pri nI all photos immediately; n fact, l
could be enough to have a contact IinIof the nedt ivcand stick the small-sca.lc
!C|CCuC!1Cn8on each folder.
Similar IUC8 will appl y in the evCnI thai you decide IO downo8d frame c1la!c
ments inyom computer (pri ntouts are still recommended IOiqut|reference) . No
matter how you collect !D!^rma|Cn,yCuUus\dCit in a W8ythat nuKCsO1-|crm
!1|0vuI 8 smooth, painless procedure. HC|e !1cs Ihe advantage O having f ranu
n!ar_cmcnt: in paper f( xm: si nce you mighI l-i nd more documentation Cn a film
tomorrow or in the distunt future, and since the photos, once t hey are pri nted,
15
o| LL|! C! N
should be net t o the relevant file, it i s best to devote a flder to each tide
yu can insert clippings, frther notes, photocopies of articles, and so on.
Even
you are intrested in silent CIDCU88Sa whole, and not just i n some particular aspec!
of it, there is no need to worry: the set of files obtained i n this way will be
respectable, but not a daunting size. And when data a.re easily accessible, you
have more time available to elaborate useful ideas, without which all this prelii
nary organisation i s nothing but a wilderness of meaningless codes, words
Du
images. Adopting and implementing a research method is only a matter of 0!-
pline; bringing the method to its creative outcome i s 8 whole difrent
matter.
Making sense of hi,story is a much larger endeavour than organising research
f,
ad the responsibility it entails is not something that caL be taught in a. research
manual. From now on, you must be your own teacher, and quite a demanding
one.
Further Readings
Despite the flourishing of silent flm studies i n recent years, l iterature on In08uD-
ject of viewing practice is still scarce. A general outine of the problems of ac.r
to flm collections can be fouJild n Documents That Move and Speak. Audioviual
Archives in the New Information Age (Munich: .. Saur, 1 992). The problem
identifing early fms has been addressed by Harold Brown, Physic.a.l Charcr
istics of Early Films as Aids to Identifcation (Brussels: FIAF,. 1 990); Suza1ne
Richard, 'Pathe, marchio di fabbrica', Segrwcinema, voL 6 no. 23, 1 986, pp. 74-7 (a
revised, French version ofthe essay was published i n >o,no. 1 0, 199 1 , pp. I
Jacques Mahhete, 'Pour une veritable archeologi.e des premieres bandes OD
matographiques', !Vb, no. Z4, June 1 998, pp. 9-2 1 . (Many more studies of
kind are needed, 1t few scholars have taken the challenge so far. ) Ethical prob
lems of research in film archives are confronted in Paolo Cherchi Usai, 'Archive of
Babel Sight and Sound, voL 59 n.o. l , 1990, pp. 48-50; Film Preservation and !t/D
Scholarship, special issue of Film History, voL no. J,Autumn 1995, and the FJF
Code of Ethics (Brussels: FIAF 1 998); Guidelines Dealing Wth Mi.use!Ue of Film
and Video 1tCh|Vc Mterial (Stockholm: FIAT/FfA, 1. 998, wth accompanying
VHS/PAL videotape). The Federation Internationale des Jth08 du Film has
published a Manual fr Access to iD L0/cltun5Dthe Jurnal of m!rc:c!vul10n,
voL Zno, ob,Decembei \. !iDII8uj, the framework adopted for the manual
is the opposite to that described in this chapter:. access is defned from the
viewpoint of the archive, 'the selection of !llms put together accDIU|u_ iIS own
agenda (programming) while 'passiv access mean that the archive _c0!Z1{
waits until an individual or a group approaches it W1|h W|ha list of requests' (p.
6) Most of the texts on flm preservation listed at the end of Chapter 3 incude a
wealth of suggestions and insights which are germane to the methodology of the
silent flm experience; sorting them nut and making good use of them is
aDClhci
question, aS the information they provide may seem haphazard and G|(contra
dictory even to the experienced. reader.
Hl STORE DE DETECTI VE 1 55
hc[D\18I1tQuOlt0!de bc_HniD_ the 8j\cI!8 t3S3 CJ(]U
CDD]H lJ 'Uiaunc OiD8I1c), 1C cinema a Ia. 8\)D8!c 1tmnnsieant, !
[atCh !!,[[ =n\hJDI0Olcj1tu DU!ID uOI0,WhCCc8ICO!Ora:DcDu
btDU_Ol |Omy attention.