Intrinsic, extrinsic, and the magic of association

By Gordon Rugg

As regular readers of this blog will know, I have an awed respect for the ability of Ancient Greek philosophers to spot a really important point, and to then produce an extremely plausible but only partially correct explanation, sending everyone else off in the wrong direction for the next couple of thousand years.

Today’s article is about one of those points, where the Ancient Greeks didn’t actually get anything wrong, but where they laid out a concept that’s only part of the story. It involves a concept that can be very useful for making sense of consumer preferences and life choices, namely the difference between intrinsic properties in the broad sense, and extrinsic properties in the broad sense.

Here’s an example. The image below shows a pair of Zippo lighters. One of them is worth a few dollars; the other is worth tens of thousands of dollars, even though it’s physically indistinguishable from the first one. Why the difference? The answer is below…

zippo banner

Image source:

In Philip K. Dick’s alternate history novel The Man in the High Castle, Nazi Germany has won World War II. One of the characters in the novel wonders whether history might have taken a different route if Roosevelt hadn’t been assassinated in 1933. He’s prompted to wonder this by the proposed sale of the Zippo lighter that was in Roosevelt’s pocket at the time of the assassination. That lighter is a unique item, worth a small fortune; the character muses about the fact that a physically indistinguishable lighter without that association would be worth only a few dollars.

There’s a useful concept in the card sorts community that handles this issue. It involves distinguishing the observable features that are part of an object from the features that are associated with the object in some way, but are not part of it.

This distinction is very useful if you’re working on the design or requirements for something. It lets you separate the features over which you have some control (e.g. the materials that you use) from the features over which you don’t have any control (e.g. a customer’s past experiences). You can go through the list of features that you’ve elicited from your market research respondents, or whoever is involved, and partition the features fairly cleanly into two sets, then focus on the set of features that you can do something about.

So far, so good.

One obvious question is what to call these two sets. This is where things start to go downhill.

In the card sorts community, the usual terms are intrinsic (for the observable features that are part of the object itself) and extrinsic (for features that are external to the object, and are not part of it in any way that can be observed in the object itself). Intrinsic is a fairly familiar term, and extrinsic is an obvious corresponding concept, even if it’s less familiar to most people.

Unfortunately, those terms have already been taken, by the Ancient Greeks, who used them in subtly different ways. There’s nothing wrong with those subtly different ways; on the contrary, they’re very useful. However, they only handle part of the issue. The ancient distinction is, in brief, between the properties of the material from which something is made, which are the intrinsic properties in this definition, and the properties of an object which was made using that material (the extrinsic properties, in this definition). It’s a useful distinction, but it doesn’t cover the same ground as the distinction made by the card sorts community.

So, there’s a need for a new pair of terms, and I’d be very grateful to anyone who could suggest suitable terms that haven’t already been used in some subtly but profoundly different way. (In case you’re wondering, the term inherent has already been taken by philosophers, again with a sensible meaning that isn’t the same as what I’m talking about here. So it goes, sometimes…)

That’s the background about this pair of concepts, by whatever name you call them. I’ll stick with the card sorts definition for the rest of this article.

Which takes us back to one of the early questions. Why should the extrinsic properties of an object be worth hugely more than the intrinsic properties, even when it’s physically impossible to tell from the object itself what its extrinsic properties are, as in the case of the Zippo lighter that was in Roosevelt’s pocket?

The magic touch…

The answer is probably to be found in the concepts of mana and of magical contagion. These are important concepts in anthropology, and central concepts in marketing, media and politics.

In brief, mana is a Polynesian concept for a sort of spiritual power. It’s not just applied to people; inanimate objects can have it. The concept is widely used in anthropology and in popular culture, and is often applied to non-Polynesian objects. For instance, objects that belonged to famous people can be described as having a lot of mana, as a result of their association with those people.

Magical contagion is a widespread concept, found in various forms around the world. The core concept is that mana can be transferred, so that a person can acquire some mana from ownership of an object. In some belief systems, simply touching the object can transfer mana; in others, mana is transferred via rightful ownership.

It’s tempting to think of these concepts as irrational and superstitious, but they’re another example of identifying a real, important phenomenon, and then coming up with a plausible but horribly misleading explanation.

A lot of the features traditionally associated with mana and with contagion magic make a lot more sense if you re-frame them in terms of concepts such as honest signalling. This is a concept from evolutionary ecology, and it describes ways of sending out signals that cannot easily be faked. Honest costly signals are widely used among animals in situations such as competition for mates, where an honest signal can help you to avoid a potentially disastrous fight against an opponent who is likely to beat you.

How do you know that a signal is actually honest? One common approach is to use costly signals, which cannot easily be faked. Yes, they’re expensive, but that’s the point; they show that you’re in a league where you can afford that expense, and they’re still cheaper than the alternative risk. Peacock tails are a classic example.

Among humans, signals can relate to a range of social currencies, as well as to financial currency. A common social currency is affiliation with a powerful person or group. If a powerful person gives you an object which is a visible signal of their trust or affection, then it’s a signal that you have powerful social support. This is why the giving of rare, high-status objects is a big deal in societies around the world. The objects themselves may be useless in practical terms (for instance, a polar bear as a royal gift is striking to look at, but not a lot of use for anything around the house). In social terms, though, they are very useful, because they’re an honest, costly signal of your status, and of your access to goods and resources that most people don’t have.

So, in summary, extrinsic features such as who gave you an object can have very real practical implications. Often, though, they are easy to fake, which is why having a solidly-established body of evidence for the provenance of the object is a huge deal in the world of antiques and collectables.

On which note, I’ll end.

Links and notes

There’s more about the theory behind this article in my latest book: Blind Spot, by Gordon Rugg with Joseph D’Agnese

You might also find our website useful for an overview of our work.

Overviews of the articles on this blog:



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s