<HTML>
<B>Please stop sending messages concerning linguistics and lexicographers
to this mail address.</B>
<BR><B>Thank you</B>
<P>Constantina Stamou a écrit:
<BLOCKQUOTE TYPE=CITE>Dear list members,
<P>this is what I received on my query about metrical scanners for English
verse:
<P>Thank you very much for your responses.
<P>Constantina
<P>From: Jason Eisner <jason@cs.jhu.edu>
<P>Not aware of anything, but such a system could be built.
<P>The first step is to find the stressed syllables when the poem is read
<BR>as prose. This might be approximated by looking up the dictionary
<BR>pronunciations of the words. In general, however, the stress
pattern
<BR>sometimes depends on part of speech ("progress," "content," etc.) and
<BR>even on context.
<P>Text-to-speech systems have to solve this problem, and you may be
<BR>able to use such a system to convert your poem to a phonological
<BR>representation that accurately marks syllables and their stresses.
<P>Now you have to get from the stress pattern to a tag indicating the
<BR>meter type. Hand-written heuristics may suffice for this, though
<BR>sloppy meter is perhaps harder to recognize than strict meter.
Another
<BR>option would be to annotate some examples with tags and train a
<BR>statistical model on them. For example, an HMM whose state records
<BR>the last few observations (stress, no-stress, line break) as well as
<BR>the hidden current tag.
<P>-cheers, jason eisner
<P>From: Susan Hockey <s.hockey@ucl.ac.uk>
<P>You can find some discussion of computers and metrical
<BR>scansion in my book Electronic Texts in the Humanities,
<BR>OUP 2000. The success of this technique depends to
<BR>a large extent on the natural language and the specific metre.
<P>Susan Hockey
<BR>*******************</BLOCKQUOTE>
</HTML>