There is a computer program for forensic linguistics, just not sure if it is available to the general public.

UKSpycatcher, I appreciate this work. A different perspective — statistics have a nice way of remaining objective. If the software that Tahoe mentions is not available to the general public (that would certainly be the faster way to go), I might be willing to take on a comparison of two letters. I am sure it is time-consuming, though, so focusing on just one is probably the best approach (to start out.)

Is there a next case you think would be worthy?

I just looked at the program I speak of is not available to the general public—not that there might be another one.

They do offer online training for those in the appropriate field: http://aliastechnology.com/training/

–See the software at the bottom of the page.

I also asked the professionals at this site about the Zodiac letters. I am curious as to their reply.

Thanks Tahoe much appreciated, it would have been much easier to feed the details into a program. It took me about 3 hrs in all. Murray- Yes it would be nice to get the data from say the Paul Stine or Melvin Belli letter. The Debut of Zodiac Letter would be best, but it would take ages. For a control ones own text could be compared.

Very interesting topic, I´m sure going to follow this work , thanks UKSpycatcher and Tahoe !

Richard, good work. I think your mentioning of the extreme use of the word, ‘I’ is telling. Zodiac was an egotist, and loved to use the word, I. "I want you to publish this" "i want you to do this" etc. Good work, I think you further prove how likely it was that Z was responsible for some of the unconfirmed writings

UKSpycatcher, I appreciate this work. A different perspective — statistics have a nice way of remaining objective. If the software that Tahoe mentions is not available to the general public (that would certainly be the faster way to go), I might be willing to take on a comparison of two letters. I am sure it is time-consuming, though, so focusing on just one is probably the best approach (to start out.)

Is there a next case you think would be worthy?

Interesting Murray you wrote 340 letters and 80 words, so at random your index was 4.25, too high to be Zodiac.

Tahoe27
I just looked at the program I speak of is not available to the general public—not that there might be another one.
They do offer online training for those in the appropriate field:
–See the software at the bottom of the page.
I also asked the professionals at this site about the Zodiac letters. I am curious as to their reply.

Tahoe 265 letters, 62 words = 4.27.

Morf13
Richard, good work. I think your mentioning of the extreme use of the word, ‘I’ is telling. Zodiac was an egotist, and loved to use the word, I. Good work, I think you further prove how likely it was that Z was responsible for some of the unconfirmed writings

Excluding Zodiac quotes and letter Z. 201 letters, 48 words = 4.19. Even being forced to use ‘I’ as a quote twice, otherwise it would have been higher. Excluding the forced ‘I’ quote 199 letters, 46 words = 4.33.
All three 4.25, 4.27 and 4.33. Zodiac 3.91 and 3.82 – average 3.86.

I can now officially clear you all as Z suspects.

Very interesting stuff.

I’ve thought before about whether some kind of stylometric approach could be used to good effect in the Z case. My main worry has always been sample size. How much text do you need to make meaningful comparisons and reach valid conclusions? The great advantage of a certain kind of purely linguistic approach is that context doesn’t matter: A writer leaves a linguistic "fingerprint" behind which can be quite unmistakable and which will betray him or her no matter what sort of text he or she produces. The problem is sample size, though. I don’t think you can meaningfully do an analysis of this sort without a certain amount of text on both sides of the comparison – but I don’t know what the minimum would be.

Looking forward to hearing more.

All three 4.25, 4.27 and 4.33. Zodiac 3.91 and 3.82 – average 3.86.

I can now officially clear you all as Z suspects.

Whew, what a relief!

Now that’s straightened out, I’ll look to focusing on one of the letters this weekend. I’m curious to see how the process goes.

Very interesting stuff.

I’ve thought before about whether some kind of stylometric approach could be used to good effect in the Z case. My main worry has always been sample size. How much text do you need to make meaningful comparisons and reach valid conclusions? The great advantage of a certain kind of purely linguistic approach is that context doesn’t matter: A writer leaves a linguistic "fingerprint" behind which can be quite unmistakable and which will betray him or her no matter what sort of text he or she produces. The problem is sample size, though. I don’t think you can meaningfully do an analysis of this sort without a certain amount of text on both sides of the comparison – but I don’t know what the minimum would be.

Looking forward to hearing more.

especially when the writer is trying to obfuscate his identity. you’re more likely to end up with mindful writings that attempt to thwart analysis. that said, maybe we’ll get lucky and z couldn’t have anticipated future achievements in computational analytics.

especially when the writer is trying to obfuscate his identity.

Indeed, that is another problem.

From the little I know about stylometry, you can fool the analyst if you actively seek to do so. But even though Z may have been trying to obfuscate and bewilder, he probably wasn’t conscious of his use of those little, insignificant words which are precisely what the analyst is looking for – the frequency of certain non-contextual, auxiliary words is what this "fingerprint" consists of.

It’s still a problem that Z’s texts could be extremely contrived in all sorts of ways. And the shorter the text, the more important it is to be aware of the nature and context of the communication. *

* Which in Z’s case includes the fact that we don’t actually know precisely what the nature and the context are.

The Zodiac certainly could have deceived the reader in his linguistic style, but one thing he would have found very difficult to disguise was his over inflated ego, just like the braggart who shows no restraint when boasting about his crimes. The Zodiac Killer could have been anybody, but his constant references of I, me, my, showed that this man was somebody who craved attention and I would be highly surprised if he had not boasted about some aspects of these crimes at some point in his life. Arthur Leigh Allen allegedly said to Ralph Spinelli that he could offer his assistance, as a form of hit-man, following this up by claiming ‘I’m Zodiac’, and stating that he could kill people and take credit for it. Ralph Spinelli rejecting Allen out of hand, was supposedly further told ‘I’ll prove to you I’m telling the truth, I’ll kill someone in San Francisco and you’ll know its me’.
I am not saying that Allen was the Zodiac Killer, but this is an example (if it were the truth) where the recipient of the claim, disregards it out of hand for being too fantastical. This is not unusual, it’s often disregarded as bravado, beer talk and people either ignore or laugh it off. The Zodiac had such an ego, he probably gave us a big clue, almost placing his own alias at the foot of the 340. I am utterly convinced the Zodiac Killer has revealed himself in some form, to somebody, but unfortunately it probably washed straight over them and the moment passed, and now its probably too late.

I am utterly convinced the Zodiac Killer has revealed himself in some form, to somebody, but unfortunately it probably washed straight over them and the moment passed, and now its probably too late.

That’s very possible. I can easily see him, say, telling someone in a manner which wouldn’t have been obvious – and then chuckling about it to himself.

I agree, of course, that his ego shows through in his letters – very much so. He’s screaming for attention in more than one way. And it is, indeed, possible that the pronoun "I" would have been over-represented in ANY text he authored, as a result of him being an egomaniac.

Context is problematic here, though. The "Confession", for instance, is just that – i.e. a personal sort of "genre" in which it is to be expected that the writer focuses on himself to a certain, if not a large, degree.

FWIW the sort of words often used for stylometric comparisons would be, for instance:

"the"
"in"
"a/an"
"that"
"but"
"though"
"however"

What they call non-contextual words, i.e. words which will appear with the same regularity (if composed by the same author) regardless of context, whether he or she is writing a novel, a letter or a scientific paper.

Ideally, that is – because sometimes the context can indeed force through a frequency (for some words) which isn’t natural, so to speak.