Ernie Lepore’s contribution to our series on the most important philosophical ideas of the century so far

During the past few decades contextualism has loomed large over philosophy. Millennia of philosophers as well as ordinary folk have apparently missed the fact that many of our most philosophically puzzling and paradox-ridden expressions are context sensitive, much like the first person pronoun “I” and the demonstrative “that”. Familiar words like “know”, “believes”, “red”, “every”, “true”, “good”, “beauty”, “tall”, even “rain” turn out to require contextualisation for application. Much as it would be infelicitous to speak of what “that” refers to outside of a context, so too “red” needs a context before it applies to anything. Fears of the Liar, the Heap, Frege Puzzles, reconciling Skepticism with Compatiblism, Moral Relativism, all go away once key expressions are recognised as context sensitive.

However valuable and commonsensical contextualism may seem, it is often missed how much it conflicts with our most cherished practices and institutions; if what we express (say, assert, claim, state, etc.) is highly context sensitive, then why do speakers invariably face no difficulty in repeating the same claim, assertion, etc. using the same words across a wide array of contexts? It is baffling how for many expressions, though there is evidence that what their uses contribute depends on features of use, there is also evidence that these distinct uses express agreement.

For most theorists, what an utterance of, for example, “Serena is really smart” says depends on a contextually salient comparison class, standards of measurement, among other such things. Yet if all you are told is that Venus uttered this, there is a sense in which you can understand what she said, viz. Serena is really smart; you can repeat what she said, i.e. say what she did; and you can indirectly report “Venus said that Serena is really smart”.

How can you achieve this without extensive knowledge of the contextually salient aspects of Venus’ original utterance; without knowing exactly which comparison class, measurement method, etc., Venus’ utterance picked out? It seems not to be a requirement on your saying what she said that your context overlap in relevant ways with hers.

Tension between alleged context sensitivity of certain expressions and their cross-contextual capacity for sharing content is not just a curiosity. Failure to appreciate content sharing reveals just how badly contemporary philosophy has lost sight of the significance of cross-contextual communication. We communicate and understand each other despite an overwhelming range of differences (in perceptual inputs, interests, cognitive processing, background assumptions, conversational contexts, goals, sense of relevance, etc.). Often, people in different contexts receive the same instructions, are bound by the same rules, laws and conventions. In order for such instructions to function, we must assume a wide range of words contribute the same content across contexts.

We hold people responsible for what they say, ask, request, claim, etc. We can do so only if we, in another context, can understand what they said (suggested, ordered, claimed, etc.), say what they said, and investigate what they said. Closely connected is that what others say often provides reasons for action. What someone said in another context can provide reason for us here only if we can understand what she said, investigate it, trust it, etc. These inter-connections and mutual dependencies between content stability and non-linguistic practices are significant because any theory that implies content isn’t, strictly speaking, shared across contexts or, at least, isn’t shared in the conversations in which we think it is, must account for the devastating implications this view has for these non-linguistic practices. To endorse a view that implies that what we do in all these cases is based on a fundamental confusion we have about the nature of our own language is a high price to pay.

Those are big picture worries; on a more local level, data in support of cross-contextual communication raise worries for a range of contextualist strategies devised in reaction to data that support context sensitivity. Contemporary philosophy and linguistics are filled with well-supported observations of instances of context sensitivity, not just comparative adjectives like “smart”, but also quantifier expressions (e.g. “every”), vague terms (e.g. “red”), semantic expressions (e.g. “true”), epistemic (e.g. “know”), moral (e.g. “good”), and psychological (e.g. “believes”) attributions. Yet these linguistic items can be used, in indirect reports, in ways that appear context insensitive. For example, suppose you’re attracted to contextualism about “know”, and so, what we say with knowledge attributions varies with context. (This point generalises in obvious ways to all the other alleged contextualist expressions.) But consider the following ordinary conversations:

• You said Igor knew that Jane was a spy.
• We have three people who, independently, concluded Igor knew that Jane was a spy.
• I’ve been considering whether Igor knows that Jane is a spy for at least two months.
• If he claimed Igor knew that Jane was a spy, then he should be fired.

We make such claims regularly. They play an important role in human interaction and in decision-making. They have vast consequences for our lives. But anyone who takes contextualism seriously is claiming to have made a most surprising discovery: we’re all wrong when we talk and think like this. Of course, we could be fundamentally mistaken about ourselves in just these ways, but at least this much is clear: if you are inclined to bite this bullet, you had better provide an alternative account of these non-linguistic practices. This fact should be at the forefront of any reflection about communication, but it hasn’t been.

The bottom line is that contextualism and our widespread practices of content sharing, at first blush, are incompatible, and so, no theory that postulates context sensitivity is acceptable unless it shows how doing so is compatible with our practices of content sharing.

Further reading
Language Turned on Itself , Ernie Lepore (Oxford University Press, 2007)

Ernie Lepore is Associate Director of the Rutgers Center for Cognitive Science

Keep up with some of the most interesting ideas around by subscribing to tpm