16:9 in English:
Bordwell on Bordwell: Part II - Functions of Film Style
By JAKOB ISAK NIELSEN
This is the second installment of an interview Jakob Isak Nielsen
carried out with David Bordwell in May 2004. This portion of the
interview concerns Bordwell's concept of style. Click
here to read Part I of the interview first.
- - -
One line of inquiry that you've mentioned here* and that seems
central to your conception of what a poetics of film could be, calls
for exceptionally close viewings of film style. You are in the process
of publishing a book on film style and stylistics.
On staging mostly. The book is on staging and is book-ended. The
four chapters in between are about one director each and then the
beginning chapter lays out some assumptions and the last chapter
responds to some criticisms. So those chapters are really about
stylistics in general.
Bodies Traced in Light - is that going to be the title of the
Figures Traced in Light, right.
It will be in line with chapter 6 in your 1997-book On the
History of Film Style but focused on the staging strategies of
four specific directors: Louis Feuillade. Kenji Mizoguchi, Theo
Angelopoulos and Hou Hsiou-Hsien?
Yes, it is an extension of that last chapter of the On the History
of Film Style-book where I try to do a small-scale example of
how one might write a stylistic history of depth-staging that avoided
some of the difficulties of traditional stylistic histories.
In your new book you propose a framework for discussing the
functions of film style in narrative films. You suggest that there
are four ways in which we talk of style in films: the denotative
function of style, the expressive, the decorative and the symbolic.
Can we flesh out these categories?
Those are just an initial shot and they can probably stand some
more refinement but just to get the issue on the table, I propose
those four. Denotative I take to be very basic but also very important:
The denotative function of style conveys things to us that we need
to know in order to understand the ensuing dramatic action. "The
character is walking into a room so we pan with the character into
the room. The character is angry, we see the character's anger and
So it is about comprehension and intelligibility?
degree, yes, it is about comprehension and intelligibility.
The second function I picked out was expressive which is - I would
say - about comprehension as well but what we want to do here is
to understand style as being used to magnify the expressive features
within the scene or to add an expressive layer to the scene. Music
is a prototypical example of the latter scenario: the camera work
may be denoting it a certain way but the music is bringing out expressive
values that are not present in the dramatic action as such. It may
be ironic for instance.
The third function is what I call the decorative function. By that
I mean that the style develops its own patterns independent of or
not necessarily arising from the dramatic action. For instance we
can imagine some camera movements that are purely functional denotatively.
That is, they just follow the actors. There are other camera movements
which are expressive such as when a character is running and we
have the character's point-of-view and the bouncy camera is trying
to convey something like the feeling of what it is like to run.
A third function is when the camera - and this we find in certain
kinds of films more than others - might start to weave its own patterns
around the action. So the camera starts by following the action
but then it assumes a high angle, goes down to a low angle, cranes
back up to a high angle and you begin to realize "well, there is
a kind of independent patterning here.This could've been filmed
differently." And it is not necessarily providing us with a better
view of the story, and it is not necessarily expressing any emotional
quality of the story but the camera is creating a dynamic pattern
of its own that we are to appreciate as simply a piece of cinematic
I of Bordwell on Bordwell.
At one level it can just be seen as making patterns.
He is interested in patterns but if you look at it from the spectator's
point-of-view this almost purely stylistic game of anticipation
and recall is part of what makes Ozu-films such rich experiences.
And that made me think about how - even within a narrative film
- stylistic patterning across the whole film can pull it itself
slightly away from the denoted function of style and take on its
The forth function is the symbolic - which I think is the most
rare - where we actually ascribe abstract significance to these
things: a large-scale thematic or conceptual point. Classic examples
would be actors' positioning with respect to figures in the space.
So for instance someone assuming a cruciform position stretching
out like Jesus. That is a clear sign that it is a piece of iconography
where we are to lock an abstract meaning: "Okay, we are to think
of this character in relationship to Christ in one way or another."
I would say a couple of things. First, the presupposition here
is that the denotative level is the most imported because all the
other functions build upon that. The second presupposition is that
any single instance of a technique, be it performance or staging
or cutting or whatever, could be performing several of these functions
simultaneously, or in quick succession. Although the decorative
is the hardest one, I think, because it depends on a bit of divergence
from the narrative functionality but in a way, any given stretch
of a cinema can be fulfilling these different functions in different
I would like for us to look at a scene from Day of Wrath
 that I think illustrates the difference between denotative
and expressive functions of style. Shot/reverse shot editing is
often seen as a relatively neutral device and you have discussed
it in denotative terms as a type of stylistic patterning that is
simply very adequate for conveying conversation scenes.
The following scene from Day of Wrath illustrates how
shot/reverse shot editing can be given an expressive twist. It is
the scene in which Absolom's mother Meret interrupts Anne and Herlufs
Marte and enters the living room just after Anne has managed to
sneak Herlufs Marte out of the room.
Let me just say a couple of things. That's exactly right I think.
And one of the things filmmakers have long known about shot/shot
reverse shot is that a couple of features are really crucial: one
is camera placement - exactly where we are in terms of the placement
of the camera - and the second is the timing of the cuts. Hitchcock
for instance often gives us a shot / reverse shot-pattern where
we concentrate much more on listeners than on speakers. The default
value is the speaker gets seen on-screen and just before the speaker
ends we cut reverse to the listener but Hitchcock, of course, is
famous for starting with the speaker, then immediately cutting
to the listener and letting us either imagine what the speaker's
expression or purpose is or letting us realize the import of the
words on the listener and maybe eliciting a growing sense of apprehension
or suspense when we realize that the listener is taking this a certain
way. So the timing of the cut and then the position of the camera
become very critical variables in short / reverse shot editing.
Exactly these variables are being brought into play in the Anne/Meret
encounter in Day of Wrath. What I want to get at is that
this particular shot/reverse shot structure takes on an expressive
function in that it is expressive of Maret's animosity towards Anne.
I guess what many scholars within the humanities would say is that
"context defines our reading of this encounter" because by
this time we already know that Meret dislikes the new young wife
of the house. But actually there are more proximate explanations
that you can flesh out when looking at the way it's shot and staged.
First, there is the brevity of the shots. Day of Wrath has
a relatively high average shot length of about 13 seconds.
Yeah, it's a rather long take film.
Now the encounter between Anne and Meret is actually cut rather
Compared to the other conversation scenes.
Yes, particularly when compared to the first
encounter between Anne and Martin. But the timing of cuts
is also significant. We see Meret enter the room (fig.1), Dreyer
then cuts to Anne (fig.2) but notice where Maret is placed when
he cuts back to her (fig.3). He has actually cut out a small slice
Yes, it's an elliptical cut.
And by doing that he makes her jump forward in the frame.
Yeah, yeah I agree, I completely agree. It is almost like an Eisenstein-cut
out of Ivan [the Terrible]
Another thing that Dreyer does with this encounter is that he
- on a couple of occasions - cuts Anne's shots off while she is
in motion and hence enhances the abruptness of the cut to the static
Meret. The cuts become much more abrupt as opposed to for instance
the Martin/Anne encounter that has Martin's voice carry over into
the shots of Anne.
Dreyer's cuts are very strange and I don't think I've done justice
to them, perhaps Edvin has [Edvin Kau, author of Dreyers Filmkunst
(1988), forthcoming in English translation]. I do agree that in
this case they have a very strict expressive function catching people
on partial gestures but I think that in terms of what he would consider
the main impression of the work which is light, space, performance,
text and things like that, these would be working at a much more
subtle level. It also becomes clearer and clearer as he goes through
his career that he thinks he can do more things with camera movements
and do them more smoothly than he can with cuts. He wants to sustain
a scene as long as possible and the cuts interrupt that. This is
one of the common strategies of long take directors. Somehow they
develop a kind of fear that the cut is interrupting or is disjunctive.
It's not smooth, it's not soft and they want that, that rhythm.
By the end, of course - in Gertrud - he'll only cut in very
specific scenes and for very particular purposes.
You know, Dreyer's a peculiar director. By the late teens and
early twenties he obviously knew the right way to make a movie by
Hollywood standards. If you look at a film like MikaŽl  it
is completely like an American or German film of the period. The
decoupage is just perfect in MikaŽl. But I think at some level he
got bored with it or thought it was secondary. Performance of course
was crucial, delivery of the text was crucial and in a way he didn't
mind things that we might even consider clumsy cuts. The way his
eye-lines work for instance. It's almost as if he begins to rethink
the scene at every cut. Instead of adhering to the 180 degree line
or a notion of the overall space, there's a way in which Dreyer
developed the idea of the ubiquitous camera -most of course in The
Passion of Joan of Arc [1928/9] but he continues it well into films
like Two People  - where he is less concerned with academic
rules of matching and, when he does cut, more concerned with treating
each figure as a solitary sculpture which he can film from the most
advantageous angle and assume that the audience will understand.
In a way that is connected to the long take camera movement aesthetic
too because if you take a single figure it is a piece of space that
I can handle in any way. Well, a long take ensemble is just another
piece of space, there just happens to be more space [laughs]. But
I can still handle in that way: I can put the camera there, make
it go around this way or that way or whatever. So I do think it
has to do with a more general aesthetic concern for a flexibility
or fluidity of how you represent space. But I agree that there are
the expressive values that you talk about at work here also, absolutely.
Any particular example is very likely. the first two anyway, the
denotative and the expressive, are often closely welded together.
So for instance performance is a borderline case. If we count
the actor's performance as part of the style of the film - and it
seems to me that we should - there are expressive values there from
the very beginning because acting or performance is letting us know
what feelingful qualities are at stake in the scene.
We only notice that they can be detached when we have inexpressive
performance like in Bresson's films where we only have a very thin
wedge of denotative information to go by. We don't know the characters'
mental states and the film's style is not enhancing them. He is
not giving us camera angles or lighting or music that would let
us in on the expressive state-of-mind of the characters. It's not
that his films are inexpressive; it's just that his means of expression
are very delicate and minimalized. So the expressive function of
style in Bresson has to be studied as a series of very brief moments
where we look and say "someone looks a certain way or closes their
eyes at a particular point in time or looks down." Those become
far more expressive than they would in a normal film.
In relation to the expressive function, you put a great deal
of emphasis on the term feelingful qualities.[see also
Figures Traced in Light, forthcoming]
In English that's an awkward phrase but it's the one that philosophers
tend to use to describe expression. You can think of the expressive
function in two ways. One, you can think of it in terms of the representation
of emotional or expressive states or you can think of .
Correct, of creating them in the spectator. The second is of interest
and importance - I'm not denying it - but in terms of style we want
to be able to pick them out. We want to be able to say of the musical
score "that sounds melancholy." Now it may not make me feel melancholy
- it may make me feel something else - but nevertheless we ought
to be able to detect these feelingful qualities in the artwork.
This is a fairly common way in which philosophers have talked about
art. In fact, many of them put expression at the center of their
conception of art and see expressive qualities as the main reasons
why art exists. That is, to put into formal design the expressive
emotional states. To me it's not the only goal of art but it is
certainly a very important one. The language we use to describe
expressive states is somewhat limited but at the same time it is
important to recognize that those expressive qualities are significant
and attached to interesting questions. For instance, are those qualities
detectable by people who don't know certain traditions of art? Or
are they immediately communicative?
I should say that all of those functions are also tied to a conception
of norms. There are norms for the representation of information
- for denotation in cinema. There are historical norms for expressive
representation, historical norms of decorative representation and
there are certainly historical norms for symbolic functions of style.
You can argue that those will vary by time and place. For instance
the way in which story was denoted in 1903 in cinema is going to
be - in many respects, maybe not all respects - different from the
way the denotative function will be fulfilled in 1950s or 1960s
cinema. And the same might go for expressive qualities. As you know,
one of the things that I've been interested in lately is the idea
of whether there are universals of artistic communication. I would
like to think that you can take a moderate position and say "well,
there are some denotative manifestations, some expressive manifestations,
some decorative manifestations, some symbolic representations which
are quasi-universal."† And there are others that are local and require
specific cultural knowledge. I don't think it has to be an all or
nothing situation. So for instance, there are very few musical universals
but one of them seems to be that increased energy is represented
by faster rhythms. No culture seems - from what I've been able to
read - to identify slow rhythms with greater energy. In a sense
it is almost a tautology. There may be very few musical universals
but there are plainly a lot of pictorial universals. Artists present
human beings across cultures and times in pretty much the same way.
They all have two eyes, they all have mouths, a nose and so forth.
At the denotative level there is a great deal of cross-over.
Which is perhaps why you are most interested in the denotative
Well, a lot of my thinking tries to be counter-current. A lot
of reckless claims have been made about representational systems
being wholly arbitrary, wholly culture-bound, "socially constructed"
- though I regard myself as a moderate constructivist. It is important
always to say the other thing: "Yes, but!" There are a great many
conventions of art, which are cross-cultural and art really couldn't
communicate if there weren't cross-cultural conventions of uptake.
We have very good empirical evidence that everyone experiences the
phi phenomenon; everyone in all cultures sees movement on the screen.
People recognize people on the screen. Furthermore, the dissemination
of audio-visual media across huge variety of cultures suggests to
us that there is a very broad, basic and robust level at which cinematic
communication can take place.
There is a more contested terrain about things like continuity
cutting. Has continuity cutting become prevalent in the world's
popular cinemas because it somehow mimics natural perception or
is continuity cutting actually a learned code that just happens
to have been imposed on the world cinemas by the strength of Hollywood
cinema?† My view is that there is a continuum where continuity cutting
is perhaps not a perfect optimum but what Herbert Simon calls "satisficing."
That is, it satisfies enough conditions of communication and of
narrative to be a flexible tool in a wide variety of traditions.
Whether it is the maximally useful tool is another story. For certain
purposes of cinematic communication and of practical filmmaking
it seems to be a very strong compromise. It allows you a great deal
more flexibility in production than a single-take scene would have
in 1910 or so; it allows for things to be adjusted in post-production;
it allows for certain kinds of audience effects that are very hard
to generate otherwise: rhythmic effects, certain kinds of suspense,
certain kinds of withholding of information, and it allows you to
time things very exactly in a way that even the most finely orchestrated
ensemble performances have a trouble doing. So it has many practical
advantages for filmmaking independent of its ability just to be
understood. But whether people had to learn it ab initio.? I'd say
"probably not." My argument is that it does piggyback on some features
of perception, which can then be generalized to text. I am not prepared
to say that this is somehow the perfect, maximally effective way
of communicating story information. It just happens to be a dominant
norm, though not a perfectly arbitrary one.
One of the questions you might ask is "well, if this is just one
style among many, how come it's lasted so long and according to
you it is going to last a long time?" I guess I would now - thinking
about it - say that classical continuity is not a stylistic school
the way Soviet montage or French impressionism or neorealism is
because all of those depend upon classical continuity. Though not
exactly parallel, it would be somewhat like perspective as a system
in painting. There is a dominant tradition historically that makes
use of classical continuity much the same way that the dominant
tradition of mainstream painting - academic painting - has used
perspective. And that's why it can last for centuries. I mean academic
perspective has lasted for 3-400 years. That would be, I guess,
my comparison with continuity stylistics. And just like perspective,
it is not an absolutely veridical representation of how we see the
world but it captures enough of the relevant regularities about
our seeing the world that it becomes a very effective and flexible
- - - - -
Part III of the interview
discusses Bordwell's own strategies for writing about film and how
he encourages others to do so. He also fleshes out how to link stylistic
inquiry to the construction of meaning and he considers consequences
of the DVD-format, Turner Classic Movies and the Internet on film
The Anne/Meret encounter - starting with the shot of Meret's entrance
- is covered in 11 shots and only takes up approximately 40 seconds
of screen time whereas the first encounter between Anne and Martin
(Anne's son-in-law) - also presented in shot/reverse shot - takes
up 75 seconds of screen time and is covered in only 8 shots.