[The fact that there are people finding the existence of the Universe (or of an Universe attuned for the apparition of intelligent beings) surprising could motivate this conceptual investigation.]
A box is labeled “Surprise”. What should it contain, in order for the content to be adequate to the label (or the label to be right)? Something which one does not expect to find in a box, perhaps. Sure, the box could contain some living creature answering to the name of ‘Surprise’, but that is besides the point.
In a provisional definition, x is a surprise IFF for any y [provided that y is a being able of using concepts, language etc. or at least of recognizing something as an x (and is doing something like opening the box containing x)], y did not think of x before being presented with x.
Now, the box cannot be empty, since one can think that the box might be empty before opening the box, so in such a case the person opening the box would not be surprised by finding it empty. The fact that anyone can reason like this doesn’t make an empty box a surprise due to an infinite regress (the box couldn’t be empty, so if it’s empty is a surprise, so the box could be empty, so if it’s empty there is no surprise, so the box couldn’t be empty a.s.o.).
But then, the box could not contain something as well, since anyone could expect the box to contain something. (On the other hand, cases in which one expects the box to contain the particular thing it does actually contain are rare – „dad promissed to give me his watch on my birthday etc.”)
However, perhaps I should have said not that x is a surprise, but that “some F” is a surprise IFF…
So, the point is that the box should contain something of an unexpected kind. Were the box to contain something which we thought wouldn’t fit in a box, that could be a surprise. But this too, is beyond the point.
Regular surprises are not like this (you open a small box and find an entire solar system inside). They are unexpected in a different sense. Think of gifts. One expects to receive some F at some point, but not at that particular point in time, so it’s a surprise. One could say: “Right, but then the moment at which the gift was received constituted the surprise, not the gift itself.”
Perhaps I am expecting to receive a wedding ring from you anytime now, but you could still surprise me by the way in which you are offering it to me etc.
So being a surprise might not be reducible to properties and relations. A surprise is something someone does (being a surprise, in the case of an object, comes from being used in surprising someone – not in any way, though; one wouldn’t call a plastic knife used by a friend to scare her a surprise; the object must perhaps be offered).
Let’s think of taking someone by surprise, then. This is supposed to be an action. So either the manner in which the action is performed is surprising or the action itself is.
The action, then, must be performed with the intention to surprise someone.
So how can we specify the content of this intention? Is it reducible to the intention to do A and the thought “Were I to perform A to P (P is the addressee, not the one acted upon), P wouldn’t expect me to perform A to P”?
One could say that this is the intention to surprise P by doing A. Of course, if this is so, then the intention to surprise P by doing an action which you have not thought of already could be expressed by replacing A with a variable ranging over the domain of actions (which can be performed by the agent etc.).
Of course, I think of the justification one would provide when asked “Why did you do A to surprise P?”. The psychological intention seems irrelevant. One could be trained to unintentionally have the neurophysiological state corresponding to the “psychological intention” whenever saluting another person (and thus to reject the question ‘why did you salute P to surprise her/him’) etc.
One could do something surprising and refuse to provide reasons for doing the action in case as a surprise. On the other hand, one could call an action “a failed attempt to surprise P” by providing the appropriate reasons. In such cases, the reasons are supposed to support the claim that “P wouldn’t have expected me to do what I planed to do” etc.
So what meaning does “expect” have in this context?
Do parents surprise their children by bringing them into life? Does the mother surprise the child by giving birth to it? If you say ‘yes’, you do not think of the right sort of expectations. The right sort of expectations must be related to some practices. When natural expectations are contradicted we say that someone startled someone else (or something similar).
But then, what about the practice of giving surprising gifts? P’s expectation is to receive a surprising gift. If S doesn’t give P a surprising gift, P is going to be surprised, but is not going to receive any surprising gift. Maybe there is no contradiction here. It is not as if P is both surprised and unsurprised (in the same sense). P is surprised by being given un unsurprising gift.
So the empty box could be used to surprise some who was taking part to the practice of giving and receiving surprising gifts.
That aside, this seems to lead to a problem about practices. Can one act to surprise someone else only by doing things, while being involved in practice R, which do not conform to R? If you conform to the practice to which you are taking part in, your actions are not going to be surprising, but the manner in which you perform them could be.
It could be said: “Either action A conforms to the practice R and then P expects you to do A, or not. If you do A in a different manner, but A still conforms to practice R, then the manner was not important, so a change in manner was not unexpected. If the manner in which A is performed is unexpected, then perhaps doing A in the manner M does not conform to the practice R”.
This, however, doesn’t seem so important anymore.
[this is a rather confused train of thoughts which I do not plan to continue at a later time]