Imagine the following scenario.
You're on trial for drunk driving. The evidence the police have against you consists in a confession which you don't remember writing and the testimony of the arresting officer. Luckily for you, the arresting officer just got fired for corruption in an unrelated manner. You plead not guilty and instruct your attorney to make the following defense.
(1) My client asserts he was not drunk the night in question.
(2) The state maintains that he was.
(3) Both of them cannot be correct. Either the state is correct, and my client was drunk, or my client is correct, and he was not drunk.
(4) An account produced while intoxicated is not sufficient to establish guilt beyond a reasonable doubt.
(5) The only potentially credible evidence the state has is my client's written confession, produced the night in question.
(6) Let us assume that the state is right. In this case, my client's written confession was an account produced while intoxicated, and is therefore not sufficient to establish the guilt of my client beyond a reasonable doubt. We conclude that my client should be acquitted.
(7) Let us assume that my client is right. In this case, my client was not intoxicated on the night in question. We conclude that my client should be acquitted.
(8) Since in both cases my client should be acquitted, we conclude that my client should be acquitted regardless of whether the state is correct or whether my client is correct.
(9) It may be objected that in (6) I have conceded that my client was intoxicated, and that therefore we can conclude that he should therefore be found guilty in this case. All this proves is that the proposition that the state is correct contradicts itself, because it makes my client simultaneously guilty and unconvictable.
This isn't a cut-and-dry case of the liar paradox, since we're not concerned with truth and falsity but the empirical establishment of a fact beyond a reasonable doubt. I don't think a jury would ever acquit someone on this basis, but I do wonder about the following.
Suppose a judge should disallow any testimony by a person who is drunk either at the time of the incident in question or at the time of the testimony, because it is so unreliable as to be irrelevant. Suppose the same argument holds as to whether a judge should allow into evidence a confession produced while drunk.
Should such a confession be allowed in all cases, or all cases where the confessor is not accused (among other things) of being intoxicated?
One probable course one might try is that in this case, the confession is always admissible, but only the form, not the content, of the confession may be considered evidence of anything when operating under the assumption that it was written while intoxicated. So supposing the writing is sloppy and there are numerous spelling or grammar mistakes (compared to something the confessing party wrote while not intoxicated), the confession would be evidence that the writer was intoxicated and only that the confessor was intoxicated.
If the confession is written in clear script or printing and is free of error, (again, compared to a writing sample made while not intoxicated), what would that imply?
Assume the confession was written while intoxicated. In this case, the admissible evidence suggests that the writer was sober. We have a contradiction, and conclude that the confession was written while sober.
Assume the confession was written while sober. In this case, the admissible evidence suggests that the confession was written while intoxicated!
It seems that under this potential ruling by the judge, we get the liar paradox.
What hasn't been considered, I think, is the bearing of the defendant's sobriety at the time the confession was written on his or her sobriety at the time they were allegedly engaging in an activity it is illegal to do when intoxicated.
If we assume that the writer is drunk, then we can only make arguments about their sobriety at the time of writing. The question becomes, if they were drunk at the time of writing, is it provable beyond a reasonable doubt that they were also drunk earlier, and at that time driving? I think that in the absence of reliable outside testimony (which is absent here) the answer is clearly no.
So we can only prove anything from the confession about the person's sobriety earlier in the evening if we assume they were sober when they wrote it. However, we are not entitled to make that assumption, as we only know that they were either sober or they were not.
I'm going to conclude that the judge should rule the confession admissable, but that when operating under the assumption that it was written while drunk it may only be used to determine whether the person writing it was drunk at the time, not whether they were drunk at any other point in the evening. I'm then going to say that in light of this, the jury should find not guilty.
So, that concludes a blog post that is neither about what I had planned to write about today (MLK day, or rather, just after midnight on the day after), which was the philosophical basis of nonviolent nonparticipation, nor really about freedom. In any case, I thought it was interesting. And probably completely wrong. Oh well.