Semiotic fallacy

The (mistaken) conflation of a symbol with the object or concept it represents.

The best-known and most impressive description of this error in thinking is probably René Magritte’s painting “La trahison des images” (The Treachery of Images).

The picture shows a tobacco pipe against a monochrome background. Underneath is written: “This is not a pipe” (in French: Ceci n’est pas une pipe).

In this context, the solution to this apparent contradiction is of course quite obvious: it is indeed not a pipe, but the picture of a pipe.

• Map-Area Relationship

Symbols are principally not the identical with the object they symbolise: The word “tree” is not a tree but a word, just as the image of a pipe is not a pipe but an image.³⁾

Even the word “pipe” is simply a linguistic convention for referring to a certain class of objects. Magritte uses the French word “pipe”, which is spelled the same as the English word, but he could just as easily have used the Portuguese “cachimbo” or the Finnish “piippu”. All these words are different symbols that – just as the image mentioned above – refer to an actual pipe. And just like the image, all these words lack important properties that only the actual object possesses – most notably the property that it can be filled with tobacco and smoked.

Even in the example above, where the distinction between the object and its image should actually be quite easy to comprehend, most viewers first tend to interpret this apparent self-contradiction as a form of surrealist aspect of the painting (which is of course encouraged by the fact that the artist is known as a representative of surrealism in painting).

It becomes even more difficult when the symbols also refer to abstract facts (e.g. at a different level of abstraction). In such cases it can be difficult even for experts to recognise this condition.

This concerns, for example, mathematical models of certain physical or even economic phenomena: these models are not the phenomena, but they are symbolic representations that seek to represent or model the actual phenomena.

The conflation can also go in the opposite direction, i.e. that a concrete phenomenon is confused with the abstract concept for it.

An example of this would be confusing the (measurable) IQ with the (abstract and unmeasurable) intelligence of a person.

The specific aspect of this fallacy lies in the confusion of symbols with the things they represent. However, there are frequent overlaps with other fallacies.

Firstly, hypostatization refers to the confusion of the abstract with the concrete. Since a symbol is also an abstraction, the semiotic fallacy can, at least in some situations, also be considered a form of hypostatization.

Secondly, Goodhart’s law describes a situation in which a metric is confused with what it measures. In some situations, this is also virtually indistinguishable from a confusion between a symbol and the object it symbolizes.

In a broader sense, all the fallacies summarized in the article on reification can also be viewed as variations of this fallacy. For example: Depending on the situation (and the definition used), the ontological fallacy can be interpreted such that the inference from the existence of a concept (i.e., a symbol) to the existence of what it describes constitutes a practical example of the semiotic fallacy.

One category of symbols that actually corresponds to the everyday use of the term “symbol” are error symbols, such as those found on washing machines or dishwashers, on a car’s dashboard, or on numerous other technical devices.

These alert the user to various problems, such as when a consumable is running low or when technical issues arise, such as clogged filters, low oil or tire pressure, etc.

These error symbols are however the symbolic representations of the issues they are depicting. They are not – at least in most cases – the issues themselves, in much the same way that a picture is not a pipe.

Without this distinction in mind, one might get the idea that it’s enough to simply clear the symbol (i.e., the error message) instead of solving the underlying problem. This leads to workarounds such as briefly turning the car’s engine off and on again to clear certain error messages⁴⁾. Or one might even consider removing sensors to prevent the error message from appearing in the first place.

It is true that there are cases where the problem actually lies solely with the error monitoring system – that is, false errors are being reported. But even in these cases, it is better to try to identify (and fix) the actual underlying error rather than rely on what may turn out to be wishful thinking.

Reporting error conditions is particularly important when it comes to modern information technology. Just like the devices mentioned above, a printer might signal that it is running out of toner, whereas a router might indicate that it was unable to establish an Internet connection. This topic also includes error messages generated by a compiler:

In software development, a programmer typically writes source code, which is then passed to a program called a “compiler.” The compiler translates this source code into the final machine code, which can then be executed by a processor. It can be particularly frustrating for beginners that, in many cases, the compiler initially outputs a seemingly endless list of error and warning messages detailing everything that is incorrect or at least in need of improvement in their program code.

A common practice is to try to find conditions under which these error messages no longer appear. Sometimes it is possible to manipulate compiler settings so that entire classes of warnings are no longer generated. One can also tinker with the source code until an error message disappears (shotgun debugging). Neither of these approaches are really useful, because the error messages are not an end in themselves, but rather an indication of structural problems in the program code.

The right approach would be to understand the underlying errors and develop strategies for resolving them. To do this, the first step would be to stop viewing the error messages as the problem and instead focus on the errors themselves.

Another example of the kind of fallacious thinking discussed here is the misinterpretation of traffic signs as if they themselves were the cause of the traffic regulation. In reality, however, they are “merely” a symbolic representation – and often a warning of – a specific traffic situation.

The real reason for a speed limit on a stretch of highway or in a built-up area is not the sign itself, but the specific traffic conditions at that location, which make a speed limit necessary or at least advisable. The underlying reasons for this can vary widely: a dangerous curve, an intersection, a high accident rate at that location, or even the need to protect residents from excessive traffic noise.

Let’s take a so-called “residential street” as a specific example: In an area marked with the appropriate signage, certain special rules apply. For example, motor vehicles may only drive at walking speed, and pedestrians and cyclists always have the right of way.

The reason to follow these rules, however, is not that there is such a sign; rather, the sign is there to warn drivers that there are likely to be pedestrians and children playing on the road, so they should drive with extra caution.

While it is certainly not a mistake to obey a speed limit because the road is signposted accordingly, it would surely make more sense to understand that you are actually doing so because you want to avoid a danger (which might not be so easy to recognize without signage). The sign is therefore “merely” the symbol that warns of the danger. The fact that it also provides specific instructions on how to deal with this danger makes it useful – but it is not the danger itself.

This section is still work in progress, further examples will be added later.

• Goodhart’s law
• Hypostatization
• Reification
• Ontological fallacy

• Map-territory relation on Wikipedia