Re: Khoisan is very old

In message <1175697237.022036.8980@xxxxxxxxxxxxxxxxxxxxxxxxxxx>, Darkstar <darkstar100@xxxxxxxx> writes
On Apr 4, 4:35 pm, Richard Herring <junk@[127.0.0.1]> wrote:
In message <1175687856.989260.106...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
Darkstar <darkstar...@xxxxxxxx> writes>On Apr 3, 11:50 am, Richard Herring <junk@[127.0.0.1]> wrote:
>> In message <1175564019.846117.287...@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
>> Darkstar <darkstar...@xxxxxxxx> writes

>> >All I'm saying is this. If a set of phenomena {A1, A2, A3} all have
>> >property S, then phenomenon A4 from the same set will also be
>> >characterized by property S.

>> That's called "guilt by association". Fallacy of accident, if you want
>> it more fornally.

[please don't quote signatures]

Okay. Noted.

>Why a fallacy? I was thinking in probabilistic terms.

Then you should state your thoughts in probabilistic terms, not as a
syllogism.

Sincerely, I don't really understand what was wrong with that
syllogism.

You said:

>> >All I'm saying is this. If a set of phenomena {A1, A2, A3} all have
>> >property S, then phenomenon A4 from the same set will also be
>> >characterized by property S.

It implies that if any three elements of a set have some property, then _all_ elements of the set also have it. That's not true unless possession of the property happens to define the set, in which case you have a circular argument instead.

> If two unique
>events coincide, it's quite probable they are connected.

Of course, it's unknown. Nothing is known. Is the cognation of the IE
family an ultimate truth? No, it's just a plausible conjecture. Maybe
I'm a creature from Mars, but it's more plausible to suppose that I'm
a bilingual from a different country. All thinking is probabilistic.

No, at this stage the probability of a causal connection is unknown.
Don't confuse correlation with causation.

I'm not confusing. I'm merely stating that correlation is *often* the
result of causation.
Consider this example from detective stories. Two similar murders
happened at Main Street. Would detectives suppose it's a mere
coincidence

Quite possibly. How many murders happen there per year?

or a work of the same hitman?

>They may not,
>but thinking otherwise has better chances of being "testable by
>experiment".

So state your hypothesis,

The thesis was that the Khoesan (or Khoisan) clicks are archaic (have
a great time depth of many thousands of years), and it's highly
improbable they're a recent innovation acquired from an unknown
source.

and describe the experiment that will test it.

So which experiments test the cognation of the IE family?

Hey, you're the one talking about "testable by experiment".

We assume
that the IE languages come from the same source because it's logical
to assume so. I can't think of any experiments which can test it.


--
Richard Herring
.