"Ed Cryer" <ed@[EMAIL PROTECTED]
> wrote in message
news:frm8t9$gaj$1@[EMAIL PROTECTED]
>
> "Evertjan." <exjxw.hannivoort@[EMAIL PROTECTED]
> wrote in message
> news:Xns9A64678A33D48eejj99@[EMAIL PROTECTED]
>> Ed Cryer wrote on 16 mrt 2008 in alt.language.latin:
>>
>>>> The problem is the somewhat crazy English habit
>>>> of separate words into smaller parts.
>>>>
>>>> Google tries to reconstruct the original,
>>>> by allowing "sub species"
>>>> to make hits on "subspecies".
>>>>
>>>> You could say it is a mistake of Google,
>>>> but the problem is in the above habit.
>>>>
>>>
>>> Google asks "Did you mean xxxxxxxxxx?" and then, usually, goes on
>>> to
>>> give hits for the string you've entered.
>>
>> Only if it suspects a mistake.
>> The above category is not in that class, they think, methinks.
>>
>>> But in this case it asks the question, and lists hits for that
>>> question.
>>
>> Sometimes it askes and corrects at the same time.
>>
>>> I figure that's wrong. It should list hits for what you've entered.
>>
>> Since when Google listens to reason in stead of to the majority of
>> users?
>>
>> Erare humanum, sponte et iniussu corrigere disputandum.
>>
>>
>> --
>> Evertjan.
>> The Netherlands.
>> (Please change the x'es to dots in my emailaddress)
>
> Well, I put in ghju iopr ty;
> and it trundled away and came up with 17 hits; although it asked;
> Did you mean: ghju iop ty.
>
> One of the hits was this site;
> http://i30www.ira.uka.de/~kutzner/allwords/allwords.html
> which contains some fascinating stuff.
>
> What analysis and linguistic conclusions has it made? Its suggestion
> appears merely to have knocked off the "r" from my "iopr".
>
> Ed
>
>
>
And then I tried Skamdangling, and got;
Your search - Skamdangling - did not match any do***ents.
Now, if your theory about Google doing some kind of John Searle's
Chinese Room look-up processing on the input parameters is correct, then
it must have concluded that Skamdangling was a good word but it just
couldn't get any hits on the Web.
Ed
|
