<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 5.50.4522.1800" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#fffff0>
<DIV><FONT face=Arial size=2>Dear Linguists,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>For my undergraduate dissertation I am wanting to
investigate questions of the 'how x' framework in written data only. To do this
I recently downloaded the BNC corpus, but as I am new to corpora, I am a little
confused. For the purpose of explaining my dilemma I will use the
example 'how big'.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Typing 'how big' into the 'phrase query' tool
yields lots of instances where 'how big' is being used other than in
questions. My solution to this was to use the 'query builder' to state that
the sentence with 'how big' had to have a question mark in it. In the 'scope
node' I stated '<text>' so the information would be extracted from written
texts only. In the 'content node' I stated (using the SGML tool) that I
wanted to look for instances where the information I wanted was occurring within
the same sentence (<s>). directly below this I put the 'phrase query'
'how big' and directly below this I put the 'word query'
'?'. When I clicked 'ok' , however, I never got passed the
'searching' stage.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>In summary then, I am wanting to know how to
investigate questions of the 'how x' framework using the BNC corpus. A
typical (perfect) example would be: 'How big was the fish that ran
away?'</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I hope you can help</FONT></DIV>
<DIV><FONT face=Arial size=2>Rebecca Rock</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV></BODY></HTML>