From Mirrors to Deixis. – Subjectivity, Biplanarity, and the Sign

Full text

Per Aage Brandt
From Mirrors to Deixis. – Subjectivity, Biplanarity, and the Sign
For Torkild Tellefsen
"There is sign production because there are empirical subjects which display labor in
order to physically produce expressions, to correlate them to content, to segment
content, and so on. But semiotics is entitled to recognize these subjects only insofar as
they manifest themselves through sign-functions, producing sign-functions, criticizing
other sign-functions and restructuring the pre-existing sign-functions. By accepting
this limit, semiotics fully avoids any risk of idealism." P. 317, the final chapter: The
subject of Semiotics. A Theory of Semiotics, 1976.
Umberto Eco has continuously felt the need to return to the basic question and to
reopen it despite its frequent closures: what is a sign / a sign function / a semiotic
event / an object of semiotic analysis? Or shorter: what is semiotics about? Recent
discussions such as Göran Sonesson’s, about the semiotic status of mirrors, in a
dictionary entry about pictorial semiotics1, made it clear to me that the question still
calls out rather loud for critical treatment. Against Eco’s view, Sonesson argues that
mirrors are indeed signs.2 This is so, he thinks, because, in Louis Hjelmslev’s
corresponding terms, two ‘planes’ can be distinguisted in the device under scrutiny: a
‘plane’ of percepts that can be said to ‘express’ a content situated in another ‘plane’.
1 The section “The mirror and the picture sign” of his entry Pictorial Semiotics, in (eds. T.A. Sebeok &
M. Danesi), Encyclopedic Dictionary of Semiotics, Berlin: De Gruyter (2010, third ed.).
2 Sonesson writes: “According to a theory first presented in Eco’s (1984) dictionary entry on the
mirror, and enlarged upon in his recent writings (1997, 1998, 1999), the mirror is no sign. In particular,
Eco quotes seven reasons for denying the sign status of the mirror, which can be summarised as
follows: 1) Instead of standing for something it stands before something (the mirror image is not
present in the absence of its referent); 2) It is causally produced by its object; 3) It is not independent of
the medium or the channel by means of which it is conveyed; 4) It cannot be used for lying; 5) It does
not establish a relationship between tokens through the intermediary of types; 6) It does not suggest a
content (or only a general one such as “human being”); 7) It cannot be interpreted further (only the
object to which it refers can).” Then Sonesson argues against each point, starting by mentioning the
views of Husserl and Piaget: “Neither Peirce nor Saussure have really defined the concept of sign, but
simply take it for granted. We can spell out what is presupposed by the sign concept by making use of
some ideas derived from Husserl and Piaget (cf. Sonesson 1992b). According to the former, the sign
requires a difference in focus and mediation. The expression is directly perceived but is not
thematic , and the content is indirectly perceived but thematic . But this criterion clearly applies to
the mirror, just as well as to the picture. Something which is comparatively more direct and less
thematic, the mirror image, stands for something which is less direct and more thematic, the object in
front of the mirror. / Piaget’s criterion depends on the notion of differentiation. Expression and content
are differentiated from the point of view of the subject. There seems to be two possible interpretations
of this conception: Differentiation may mean that the expression does not continuously go over into the
content in time and / or space; or that expression and content are conceived as being of different nature.
In both senses, the mirror is certainly as sign. The person or thing in front of the mirror is clearly
differentiated from the image in the mirror.”
1

The term ‘expression’ is taken to mean that the perceiving subject apperceives
something through something else. We might call this principle an experienced, or
phenomenological, biplanarity. There is a sign when there is biplanarity in the
phenomenon. Therefore Peirce’s index is a sign; e.g., on one plane, proximal, smoke;
on the other, distal, fire. Additionnally, a mirror is an artefact, made with the purpose
of expanding the field of vision. In my own terminology, mirrors pertain the the
category of probes: microscopes, telescopes, weathercocks, sundials, termometers,
etc., are all artefacts that allow us to extend our field of perception and obtain
knowledge about parts of the world that otherwise would be out of perceptual reach.
So the experiences they offer are ‘biplanary’ and would be signs in this sense. The
probes themselves may be signs in some sense, indices of human curiosity or
something of that kind, and the apperceptions they offer are indeed signs if we accept
the clear and simple definition by biplanarity.
However, I would firmly object to that view, both by sharing the intuitions that
made Eco reject the determination of mirrors (or mirror experiences) as signs, and by
adducing supplementary arguments from cognitive science. In fact, sheer biplanarity
is rooted in human cognition. This is why the logician C.S. Peirce was interested in
the first place. Seeing things as manifestations of other things is the general condition
for thinking: establishing connections – sometimes through logic – between percepts
and concepts. We do this in every second of our waking life, and we even try to do so
when we assign meanings to our dreams. More specifically, our minds are able to
perceive sounds and visual events, and even other sensory events, as representations
of states of affairs situated outside of their own immediate time and place frame, as
imaginary events or states. These sensory events can be cognized as representations of
virtual events or states in at least two ways: as imagery or as signaling. The mirror
image offers imagery and most often also signals to us some critical state of affairs
(examples: my own face; shaving needed!; rear-view in car: other car approaching;
watch out!). These two representative functions, imagery and signaling, can evidently
be imitated and inscribed in the realm of human communication, and they may then
be called iconicity and symbolicity, respectively. But the biplanarity of such
representations does not make them signs, unless we absurdly decided to identify all
perception as semiotic. What makes biplanary events semiotic, and thus makes their
bindings between percept and concept become sign functions, and not just cognitive
2

functions, is their communicative function. We can, with our body, produce images
and calls. We have developed this capacity into language. The linguistic sign is not
just a structured biplanary phenomenon, as Hjelmslev thought3, but an iconico-
symbolic routine in human communication, which builds on cognition4 and in
particular on the mechanisms of shared attention, or thinking-together, by blending
other subjects’ expressed thinking into individual processes of making sense. This
perspective changes the discussion on what constitutes the sign, not only concerning
language.
The percept is really an expression, according to this communicative-cognitive
view, if it is intended by a subject to mean something outside itself to another subject
(or to the same subject at another moment), that is, if it expresses a subject’s
communicative intent and a content intended to be decoded and understood by
another subject. In typical cases, the expression is itself a percept produced
intentionnally by the subject, now to be termed a subject of enunciation, or
‘enunciator’, and intentionnally addressing a subject, an ‘enunciatee’, variably
specified. We may say that communication as such consists in using the biplanarity
built into human cognition in order to share the thinking which is based on it. S1
addresses a percept (grounded in a certain situational context) to some S2 in order to
have S2 perceive it as a representation of a ‘meant’ content, a meaning, a concept
meant by S1 to be shared in that context. Such an idea is of course incompatible with
structuralist impersonality (and ‘idealism’ in Eco’s sense), since it involves embodied
consciousness, attention, intention, intent, volitive and communicative activity;
therefore, a host of mysterious impersonal formulae have been proposed when the
sign function had to be defined. Unless it was left undefined.
Our initial quote from Eco’s treatise must introduce the subject, not as a mere
‘logical’ subject of an ‘inner’ experience of biplanarity, as in Husserl or Piaget, but,
Eco writes, as an instance “producing”, “criticizing”, “restructuring” sign functions, in
order to obtain a definition of the sign function, that is, the function that makes the
3 See P. Aa. Brandt, “From Linguistics to Semiotics. Or: Hjelmslev’s Fortunate Error”, at:
https://www.academia.edu/6819523/From_linguistics_to_s
%C3%A9miotics._Hjelmslevs_fortunate_error
4 See P. Aa. Brandt, “Linguistic Theory in the Framework of a Cognitive Semiotics:
The role of Semio-syntax”, in (ed.) P. Konderak, Establishing Cognitive Semiotics, 2016 (forthcoming).
3

relation between the two planes of cognitive double-perception a semiosis. Eco tries
to avoid the structural impersonality of a totally non-subjective sign function, but
without venturing into explicitly communicative lands. Still, his mirror is not a sign
despite its biplanarity. What you see in the mirror is not ‘in the mirror’ but ‘out there’
where the optical source of the image is situated. The mirror image is not an
expression in the communicative sense, since it has no enunciator.
In a recent study5, I have proposed a general model of the sign function that
includes the necessary enunciative structure and thus accounts for its cognitive
grounding. The simplest version of a sign is taken to be a deictic gesture.
Fig. 1. A simple deictic gesture: pointing. [Anonymous photo].
In fig. 1, a S1 addresses a S2, us, by eye contact (and, supposedly, by previous verbal
contact) and points towards an item (O) on the wall, in casu a photo. The triad S1-S2-
O is constitutive of this relation. Here is a famous pictorial example of pointing:
Fig. 2. Nicolas Poussin, Et in Arcadia ego.
5 “Deixis – a Semiotic Mystery: Enunciation and Reference”, to appear in the journal Cognitive
Semiotics, 2016.
4

The general structure of enunciation, whether considering dialogue or reported
discourse, is the generative cell: S1 shows S2 an O. Showing is a form of giving
(giving-to-see, French: donner-à-voir). In dialogue, S2 then becomes an instance of
S1, and S1 is universally supposed to have been an S2 (enunciatee) to some other
instance that informs his present enunciation as a source of knowledge, or authority
(A).
But in fact, in all cases6 of deixis, S1 shows S2 a signifier, an expression,
which constitutes the immediate object (O) of shared attention. In turn, this signifier
will show S2 a content, a signified. The signifier is now itself an enunciator, the
representative of S1 that shows and signifies the concept, the content, the signified, to
S2. So the intersubjective relation S1-S2 is directly involved in the sign function
relating signifier (sa) and signified (sé). This is the new, enunciation-based view of
the sign function that I wanted to introduce. The sign is based on deixis (S1-S2-sa &
6 I am really saying that in all cases of deixis, the O is a signifier! If you point to a present object, e.g.
in a baker’s shop: “How much is this [or that] bread, please?” – the object can be materialized by a
different loaf in the shop, so it serves as an image of what you ask about. – You hear a metallic sound
afar: “Was that a phone call?”
5

sa-S2-Sé). In order for S2 to ‘decode’ the signifier, some knowledge of what
motivates S1’s act of signifying is required. The model proposes a pre-present
instance (A-S2-R) that lets some authority inform S2 (to become S1 in the present)
about a contextual state of affairs (R) (a referential grounding in reality) which
explains how the signifier can mean the signified. We always speak ‘from
somewhere’, as Michel Foucault said. I add that the addressee is supposed to know
and presuppose this, and, in general, to automatically guess which implicit context is
likely to be the relevant one, unless the enunciator presents himself explicitly as a
representative of a named A addressing a circumstance of a named R.
If this is correct, the sign function is really a deictic function of enunciation
that posits the sign – sa/sé – as an instance of communicating intersubjectivity,
without which there would be no connection between the two ‘planes’ of the sign. If I
am right, these are distinct ‘planes’ of enunciation, not of a preset impersonal function
of immanent interdependence. Impersonal sign systems only exist as fossilized parts
of living signification. Even in the most rigid sign systems, such as those of the traffic
code, enunciation posits the planes institutionally, which means intersubjectively.
Fig. 3. The enunciative structure of deixis, creating the sign function.
This new view of the constitutive relation between deixis, enunciation, and
sign function changes the semiotic landscape considerably. All expressed signs are
deictic. The object of the triad S1-S2-O is really a signifier supposed to ‘mean’ a
signified to S2. How this sa-sé relation obtains, or is ‘coded’, depends on the A-S(2-
>1)-R stance, and here is where ordinary sign classification starts. In fact, both the
signifier and the signified must already be cognitively biplanary. The signifier percept
6

must be a token of a type, which makes it iconic: an image of some type of percept; a
sound sequence, for example, may be the token of a type that is a word in some
language, or a phrase in some music. The signified is a concept activated by that word
or phrase, presupposing that language or music, and has to unfold a semio-pragmatic
value in the context of speech or playing, so its biplanarity is symbolic: its concept
represents a personalized semio-pragmatic message. This will be clear in the
following example.
Let us consider a public monument.
Fig. 4. A French World War I memorial monument.
The column is in itself deictic, like a raised finger or a tomb stone. The statue and the
cock are pieces of imagery that together make up the compound concept /French
soldier/. The list of proper names on the sides of the column will then be read as
signified by this iconic compound, and will prompt for the understanding that these
names refer to people from this place who died as soldiers during the war in question.
The deictic formation thus contains an iconic biplanarity that comes to signify a
symbolic biplanarity: remember them and honour their deed! The semiotic result is a
structure slightly more complex than those offered by Saussure’s or Peirce’s basic
7

models. The internal biplanarities were what Hjelmslev mistook for form-substance
relations on both planes. They account for the fact that no sign can be said to be
exclusively iconic or symbolic; all signs are both, in variable ways.
Fig. 5. The components of the sign function.
The semiotic biplanarity differs clearly from the purely cognitive biplanarities in that
the latter do not rely on any enunciative grounding, since they stem from ‘pre-
semiotic’ cognitive devices available and active in the human mind. Their integration
in the semiotic function, by contrast, is only possible because the enunciative
grounding in deixis creates a relation of signification between them. In life, we find
this configuration ready at hand (or rather ‘at mind’) and forget the role of deictic
intersubjectivity at the root of semiosis, or we add it as an external circumstance, as in
Roman Jakobson’s famous six-point communication model. But as Umberto Eco
notices (in the quote above), it is essential to semiosis that subjects not only use it but
also produce, criticize, and change it. These semiotic activities are possible and
occurring because subjectivity is internal to the sign function. There is no subject-free
immanence.
8