On this blog, I think it’s important occasionally to type up the material that informs my perspective on things — the ideas that feel pretty basic but never fail to surprise me with their immense depth when I sit down and consider them for a few minutes. One such idea forms a cornerstone of Peircean semiotics, and it’s there are three basic types of signs: the icon, the index, and the symbol. And the icon is the most fundamental and primitive of the signs we encounter — it’s impossible to shake it from how we engage with the world.
I. Icons (or iconic signs): what they are
Let me break these three types of signs down real quick, for the uninitiated, before we focus on the icon. According to the Wikipedia article on this topic, the “symbol” is
a sign that denotes its object solely by virtue of the fact that it will be interpreted to do so. The symbol consists in a natural or conventional or logical rule, norm, or habit, a habit that lacks (or has shed) dependence on the symbolic sign's having a resemblance or real connection to the denoted object.
For Peirce, the symbol was the purest and most advanced kind of sign there is, for it is the basis of all language, and thus the basis for all of human thought — or at least conscious, discursive, not-merely-instinctive “online” thinking. He believed that humans cannot engage in thought without language, implying that mentalese, i.e. the idea that humans can think without a language (which has been promoted by Jerry Fodor and Steven Pinker), is a fiction. Thus, the two other signs, indices and icons, are in some sense degenerated symbols even though the entirety of the animal kingdom recognizes those two and can only faintly acknowledge symbols, if at all.
The “index” is the second type of sign, and here’s how Thomas Sebeok describes it in his Signs: An Introduction to Semiotics, 2nd ed. (2001):
An index is a sign that refers to something or someone in terms of its existence or location in time or space, or in relation to some thing or someone else. Smoke is an index of fire pointing out where the fire is; a cough is an index of a cold; and so on. These signs do not resemble their referents, like icons; they indicate or show where they are.
Here we’ve got a primitive type of sign that animals can recognize, and it comes closer to human language, but it’s not quite there. When you tell your dog, “Time for supper!” and he wags his tail, it isn’t because he can comprehend the sentence at a syntactical level: it’s because the word “supper” tells him that food will arrive in the imminent future. Those famous experiments of Pavlov and his dogs involve indexical signs. Similarly, parts of our very own human grammar have indexical qualities. Deictic markers that indicate some nondescript direction, person, or time, like “there,” “then,” “here,” “now,” “later,” “before,” “you,” “me,” “they,” etc. take on an indexical function. I believe that the argument of Terrence Deacon — namely that early humans probably used words indexically before they started to use them symbolically — is correct.
But the linguistic property I’d like to focus on in this discussion is the icon. And the icon is, according to Sebeok,
a sign that is made to resemb1e, simulate, or reproduce its referent in some way. Photographs may be iconic signs because they can be seen to reproduce their referents in a visual way. Onomatopoeic words are also iconic signs because they simulate their referents in an acoustic way. Commercially produced perfumes that are suggestive of certain natural scents are likewise iconic, because they simulate the scents in an artificial way. The list could go on and on.
I believe the icon is particularly important because it’s the most basic form of sign, it informs the semiotic process of all animals, and it is impossible for us to overcome its importance. The iconic sign was probably the first means through which early humans communicated with each other. The oldest known manuport, i.e. an object taken from its original location and moved somewhere, is a tiny little thing called the Makapansgat Pebble. It was probably moved to a cave about 3,000,000 years ago, and it happens to resemble a human face. In other words, some early human from long before man reached the status of homo erectus picked up this thing because he thought it looked like his own face and decided to move it somewhere.
We humans have been in the “found art” business for some time, and it has been an activity rooted in iconicity.
However, plenty of self-styled intellectuals nevertheless like to fantasize about how we’re living in a world in which all of the signs we encounter on a daily basis lack a motivated connection to what they purportedly signify. The notion of the iconic sign offends them. They want the icon to simply go away and leave us alone forever, particularly when it attains visual form. Let’s find out why.
II. The 1960s criticism of iconicity
Discussions on iconic signs have been going on for centuries, as the concept pertains to basic matters of philosophy, but there is one recent historical example of how intellectuals turned against the idea of iconic signification: in the late 1960s there was a major debate about iconicity, and particularly whether icons were “natural” or “conventional,” that is, motivated by nature or encoded by agreed-upon conventions that human societies had developed (see this article for a full overview). There was a huge backlash against the notion that there could be such a thing as a sign that immediately makes sense to its observer without some sort of cultural mediation. And the reason for this backlash was that left-wing academics, particularly in Europe, wanted to use semiotics strictly for the study of culture and create as much doubt as possible in the idea that human culture is perfectly “natural” and therefore (so the implication would go) just in its current manifestation. Umberto Eco, who played a big role in this debate, put the crux of the issue succinctly when he said that the whole aim of semiotics should be to “possibly reduce the facts of nature to phenomena of culture, and not try to trace facts of culture back to phenomena of nature.” There was a bias built into the approach to the field, even though it was not present in the philosophy of C.S. Peirce.
The biggest target of this tendency was the human visual sense. At no point in the history of intellectual discourse has the sense of sight taken more abuse than the 1950s-70s, even as the variety of attacks came from numerous angles. Marshall McLuhan was critical of print culture for imposing an excessive reliance on the visual sense at the expense of all others. Marxists were critical of the “society of the spectacle” for helping to impose a false consciousness on the potentially revolutionary subject. Poststructuralists advocated for an ethics of blindness. Nonconformist protestants like Jacques Ellul were disturbed by the “humiliation of the word” at the hands of the truth-obscuring visual element.1
One major figure that semioticians recruited to put the notion of visual iconicity into question was the art historian E.H. Gombrich, who developed a conventionalist view of visual realism in his Art & Illusion (1961). Eco quoted Gombrich’s story of how the medieval artist Villard de Honnecourt saw a lion in real life, went home, and then drew it from memory, but his drawing relied heavily on visual clichés (or “schemata”) that you’d find in other medieval drawings used to depict other animals, ultimately causing the picture of the lion to look hardly at all like how we’d picture one today. Eco’s point, following Gombrich, is that these visual clichés, or what he called “codes,” actually influenced the way Villard saw the lion in his mind’s eye, and so we too have our own culturally-transmitted codes that influence how we internalize what we see.
I’m of the opinion that Gombrich occasionally took his argument a bit too far into pure conventionalism, but this and other points he raised were still pretty good. In fact, it was ultimately productive for the concept of iconicity to take the beating that it did, since it left us with plenty of ways to identify ruptures and disjunctions between an object and its representation through various media. But it didn’t at all eliminate the importance of the icon, nor did any of these scholarly interjections ever justify the view that semiotics always should be about resisting the urge to trace what’s in culture back to the facts of nature.
It’s also worth noting that even while Gombrich was challenging the idea of “realism” in the visual arts, he himself had produced material that absolutely affirms the idea of iconicity while de-stressing its visual significance. In his essay “Meditations on a Hobby Horse; or, The Roots of Artistic Form” (1951), he talks about hobby horses — you know, those little wooden sticks that kids ride around like a horsey. There’s nothing visual about them that invites a comparison to a horse. Instead, what forges the connection is each one’s rideability, i.e. the way the human body wants to interact with the object in three-dimensional space. Though Gombrich never uses the word, we’re now knee-deep in Jakob von Uexküll’s concept of the Umwelt, or the subjective phenomenal world that each animal (including the human) inhabits. Thus, cats and dogs will chase certain objects like neon orange balls, thinking of them as prey despite a total lack of visual similarity, simply because they’re chaseable.
But make no mistake: nothing about the Umwelt precludes the importance of the visual element. It’s just that it isn’t necessarily visual; a sign’s iconicity depends on its total cumulative effect on the subject. Over in Australia, for example, the giant brown jewel beetle, Julodimorpha bakewelli, has nearly gone extinct because it keeps trying to have sex with a certain kind of brown beer bottle — probably because the bottle looks like some kind of hyper-eroticized, ultra-fuckable female. And, again, its visual appearance counts. They certainly don’t do this with green-colored bottles!
III. Iconicity and cognitive types
As for Umberto Eco’s early criticisms of the naive acceptance of iconicity… the natter is somewhat complicated. There’s no question that Eco was motivated by antifascist politics back when he took up semiotics in the 1960s, and he has often given too much credit to fashionable left-wing intellectuals, including various poststructuralists whose work is frankly inferior to his, while also being too dismissive of genuinely interesting if unconventional thinkers like the aforementioned Marshall McLuhan. But his main advantage has been his genuine scholarly concern for figuring out the truth, and so he has returned to the concept of iconicity numerous times throughout his career. By the 1990s, he had formed some ideas that move closer to what C.S. Peirce had originally thought when he first wrote of iconicity, and he shifted his emphasis from the conventionality of iconic signs to the way the contact with the real world can modify what he calls a “Cognitive Type,” which is a model through which the mind organizes information about a phenomenon it has encountered, determining what’s relevant and irrelevant to its sphere of interests. Here’s him showing how it works while imagining what the Aztecs were thinking when they saw the horses of the conquistadors for the first time (from his Kant and the Platypus, 1999):
Of course, after having seen some horses, the Aztecs must have constructed a morphological schema not that dissimilar to a 3-D model, and it is on this basis that the coherence of their perceptual acts must have been established. But by speaking of a CT [cognitive type], I do not mean just a sort of image, a series of morphological or motor characteristics (the animal trots, gallops, rears); they had perceived the characteristic neigh, and perhaps the smell, of horses. Apart from the appearance, the Aztecs must have immediately attributed a characteristic of “animality” to the horse, given that the term maçatl was immediately applied, as well as the capacity to inspire terror and the functional characteristic of being “rideable,” since it was usually seen with human beings on its back. In short, let’s say that the CT of the horse was of a multimedial nature right from the start.
Here the concern, while not outright contradicting any of his previous work, is now much more focused on the raw experience of novel information rather than the manner in which visual information is culturally mediated.
Because of the strong political motivations of 1960s semiotics, which prompted the intellectual desire to put the very notion of iconic resemblance into question as strongly and often as possible, the concept grew more and more refined over time. Rather than naively saying “this resembles that,” “that resembles this,” semioticians were forced to consider what exactly it means to experience resemblance itself, and even ask questions like “resemblance to what? The thing itself, or merely the way we’ve experienced it according to what our memories tell us?” So essentially, semiotics has had to go back into the territory of Descartes and Kant just to make sense of the concept. As Sebeok puts it,
Contemplation of the icon sooner or later tends to turn from legitimately semiotic concerns, in the technical sense, to intractable, indeed mind-boggling, philosophical problems of identity, analogy, resemblance, and contrast, similarity and dissimilarity, arbitrariness and motivation, geometry and topology, nature and culture, space and time, life and death. The experience is like entering a fun house furnished with specular reflections and distorting mirrors, doubles and replicas, emphatic stimuli and superoptimal models, and being taken for a ride in the clairobscure on one of Gombrich's pedigreed hobby-horses.
But no matter what, there is no question that the icon is a valid type of sign, one that ought to be kept in mind at all times when doing any kind of semiotic analysis, cultural or otherwise. In fact, it’s required for us to have any other kind of sign! I’ll give a quick example. Look at these two pictures next to each other:
How do I know these two separate configurations of black lines — both symbolic signs in that they belong to language — are creating the same word, even though they both look different? There is no answer other than iconic resemblance. True, there are some culturally transmitted codes that have informed how I determine the resemblance (like how I’ve been taught the differences between cursive and script writing, or how I’ve been able to identify two different words by virtue of spaces resting between them), but without that basic fact of iconicity, I would have no ability to determine any kind of resemblance between one word and any other instance of the same word.
The other thing worth bearing in mind about iconicity is that there needn’t necessarily be an ur-type against which we compare all subsequent representations when we determine what something resembles, like a Platonic form or something like this. There may be some kind of naturally occurring ur-type for actual kittens (more on that in the next section), but there is no naturally occurring ur-type for the word “kitten,” which is purely conventional.
Because there isn’t necessarily a Platonic “form” for iconic signs, they can easily flip around and themselves become the referent. There isn’t necessarily a one-way process with iconicity. For instance, let’s say I become obsessed with some celebrity whom I’ve seen on video a thousand times, yet never once in real life. If I see her in real life, I’m at some point going to be focusing on the extent to which she compares to the videos I’ve seen — probably at first sight, if we’re being honest. The videos will become the referent, and the real, physical person, the iconic sign… even though my whole life I had taken those videos to be signs of a real person who exists “out there” somewhere.
IV. Archetypes and inherited memories
I want to conclude with one idea I’ve been thinking about for quite some time, and it has to do with the notion of the “archetype,” or the tendency for certain images to recur over and over again in poetry, narratives, and visual art forms over vast expanses of time, cross-culturally. As I stated in the last paragraph, not every iconic sign requires an ur-type, like a Platonic form. Yet nonetheless, we do have instincts that motivate us to seek out certain phenomena and make use of them, and this principle applies to all species. How does a spider, for instance, know where to build a web? It can’t be purely based on sight, since spiders have generally terrible eyesight and can distinguish little beyond light and dark. The spider has to detect something like a slant on an object; it has to experience something in three-dimensional space that lets it know, “Yes, here’s a good place.” It has to seek out something that conforms more or less to an inward type that it can identify entirely by nature rather than education. The same is true of a beaver, seeking out wooden logs and other objects with which to build a dam.
We know that these animals don’t behave in these ways due to training or mimetic observation. They behave as they do because of some natural inheritance that simply makes them do as they do and seek out various things (i.e., types of surfaces, objects, terrain) without thinking, and act upon them. These habits die hard, too. One Russian scientist produced domesticated foxes after about forty generations of selective breeding, and despite them being quite tame and friendly, those foxes would still dig holes in various strategic places in the same manner as their forebears would have done. When these animals find logs for a dam, slanted corners on which to build webs, or wet grass suitable for hole-digging, it seems that they must identify these things iconically, with each object resembling a loose approximation of what to look for that resides entirely within the animal.
So having acknowledged all that, it shouldn’t be hard to acknowledge why so many people find plausible the theory that humans, too, have a collection of inborn approximations or “archetypes” that form referents for certain objects, images, and other phenomena they find in real life. If so many cultures find the serpent, for instance, to be a symbol of evil, perhaps there’s some inward pre-cognitive schema that each man contains within himself that each serpent more or less conforms to. I myself believe there’s something to this idea. I believe we seek out objects that iconically resemble loose schemata that reside within us a priori and thus can’t be called “cognitive types,” since cognition has no involvement in the process.
This view is, at least, the assumption that drove a few different schools of thought during the early to mid-20th century. The most famous of these was C.G. Jung’s post-Freudian psychology, which has always maintained a popular appeal, and which recently became newly famous due to being so influential on the famous psychologist and self-help book author Jordan Peterson. The second most famous was that of René Guénon, who preferred to examine the phenomenon from a purely religious perspective. Whereas Jung saw these archetypes as expressions of ideas buried within a collectively held consciousness, Guénon basically felt that they came directly from God and contain the key to man achieving inward transformation, reaching a higher state of being.
The main problem with these two expressions of an “archetype theory” (they were often called just “symbols,” a word I won’t use since doing so will add nothing but confusion) is that the archetypes are often specific and fixed in a way that can’t withstand scrutiny. The problem that Marshall McLuhan always had with this kind of thinking — which he attacked by proxy by criticizing Mircea Eliade and Northrop Frye — is that it completely ignores the “ground” through which each archetype is conveyed, and for McLuhan, the “ground” was the medium of information transmission, be it oral expression, the written word, radio, video, or whatever else.2 This is generally a good point. The linguist Calvert Watkins made a similar point when he talked about the cross-cultural analysis of oral-mnemonic formulae in epic poetry: an image from a given formula doesn’t really mean anything if you’re ignoring the actual poetic structure through which it’s conveyed.
For that reason — and I’ll stress here that this last point is fairly tentative and uncertain — I’ve long been interested in going deeper than Jung and Guénon, and I suspect the two guys to really grapple with are instead Aby Warburg and Ludwig Klages. Warburg, an art historian, and Klages, a psychologist and philosopher, both used the careful study of images to suggest that various frequently recurring artistic subjects must have some kind of universal significance approaching the Jungian “archetype” (they both wrote before Jung came up with his theory). What makes them more interesting than Jung, however, is that they never committed their studies to hard and fast conclusions in which one thing symbolizes this, another thing symbolizes that, and so on. In other words, they didn’t set up a system of precise correspondences like Jung and Guénon or any of their followers. They each simply provided loose sketches for what would later become “archetype theory” without necessarily overstressing anything, keeping each possibility rough-hewn, multidirectional, and, importantly, dynamic (Warburg in particular was obsessed with the idea of objects in motion and always was remarking upon the visual suggestion of motion in Renaissance art).
The problem with both is that they privileged the image far too much at the expense of all other senses, and thus in an indirect way resemble the kind of thinking that continental philosophy attacked during the 50s-70s. Nonetheless, I think it could perhaps be fruitful to revisit the thought of each and see if it’s possible to take what they observed and expand it to a more comprehensive analysis of culture, one that takes into consideration a) a phenomenon’s impact on all senses rather than just sight, b) the medium through which the information (of whatever kind) is conveyed, and c) the experience of a cultural phenomenon as it reaches an observing subject in three-dimensional space, whether he’s looking at something through a screen on his cell phone or seeing it happen before his eyes in real life.
Until then, however, I’ll have to put a bookmark in this discussion. I’ll probably come back to it later.
Although I haven’t yet read it, Martin Jay’s book Downcast Eyes (1993) seems like a promising overview of this intellectual tendency
Though it’s good criticism, I’m not convinced it applies to all the archetype people. For instance, Guénon actually did state that sacred archetypes can become corrupted when the conditions through which they’re presented shift. For him, there’s nothing inherently Satanic about the upside-down cross, but because of a concerted effort to distinguish it from the right-side up one, it became a Satanic inversion simply due to the outside mediation.