Re: Larry Athy and His Logical Analyses.

"Eric Stevens" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> On Mon, 24 Nov 2003 10:30:34 -0000, "JMB" <[EMAIL PROTECTED]> wrote:
>
> >"Eric Stevens" <[EMAIL PROTECTED]> wrote in message
> >news:[EMAIL PROTECTED]
> >> On Sun, 23 Nov 2003 21:20:50 -0000, "JMB" <[EMAIL PROTECTED]> wrote:
> >>
> >> >> "Eric Stevens" <[EMAIL PROTECTED]> wrote in message
> >> >> news:[EMAIL PROTECTED]
> >> >
> >> >--SNIP--
> >> >
> >> >>
> >> >> >
> >> >> > Individual languages have characteristic letter frequencies.
> >> >> > http://www.bckelk.uklinux.net/words/etaoin.html shows that letter
> >> >> > rankings differ between languages. Soo must frequencies.
> >> >> >
> >> >> > Also, so too do individual styles of writing.
> >> >> > http://deafandblind.com/word_frequency.htm
> >> >> >
> >> >> > See http://www.chass.utoronto.ca/epc/chwp/tirven/tirv2.htm which
uses
> >> >> > Pearson and Chi-squared for comparing texts on a word basis.
> >> >> >
> >> >> > http://www.ils.unc.edu/~mornj/voy/voy_findings.htm statistical
> >> >> > analysis (including Pearson) is used for letter frequency
analysis.
> >> >> >
> >> >> > Such analyses are rife throughout the field of linguistics. It is
a
> >> >> > trivial exercise to compare two bodies of text to help determine
> >> >> > whether or not they are similar.
> >> >>
> >> >
> >> >Just to point out also that not one of those links is to a site that
is
> >> >using the Pearson coefficient and t-distribution in order to show a
link
> >(or
> >> >not) between two languages, which is what Larry used those techniques
> >for.
> >> >In fact, the t-distribution wasn't even mentioned on any of the sites,
> >and
> >> >Pearson was used for a completely different type of study.
> >> >
> >> I have to ask whether you are being stubborn or require to spoon fed?
> >>
> >> What Larry did is entirely consistent with the context of the above
> >> (and many similar) URLs above. Whether or not he did it correctly is
> >> another matter.
> >
> >It is not consistent with the context of the above links.  Larry used a
> >combination of methods that are not used in any of the above links to do
a
> >study comparing two different languages which is also not the type of
study
> >being used in any of those links.  You have found a few examples of some
> >statistics being used in a linguistic context, you have not shown an
example
> >of the statistical analyses that Larry used in a linguistic context, nor
> >have you shown any statistical analyses being used in the same linguistic
> >context as Larry was doing.
> >
>
> I continue to be amazed at the way you seem to take exception to
> thoughts and ideas for which you do not know of an exact precedent.
> Statistics is a mathematical tool to be used for extracting
> information from raw data. The thinking part of the brain is a tool.
> Larry used both statistics and his brain to try an idea which seems to
> me to be perfectly obvious.

It seems perfectly obvious to you here, yet in other places you have
expressed a doubt as to whether or not the t-distribution should have been
used.  Are you just arguing for the sake of arguing?  It certainly sees that
way.

>
> It seems to me that if you were in a strange building you would not
> open a door and explore whatever was on the other side, until someone
> had shown you it was safe and proper to do so. On the other hand,
> Larry (and I) would try the door to see what what was behind it.

No, we'd all try to open the door, I would just want to know that the key
fit the lock first.

>
> In this respect, I cannot really understand what you are making a fuss
> about.

I have asked Larry why he selected those particular analyses, and why he
thinks they are valid for what he has done..  You yourself have expressed
doubts in this matter, so why are you getting so upset when I ask Larry to
give somme details?

>
>
>
> Eric Stevens

-- 
John Byrne
www.iol.ie/~archaeology
To email me use the feedback form on the website.
The address attached to this post is just a spam trap.