Received-SPF: pass (google.com: domain of ilmen.pokebip@gmail.com designates 2a00:1450:400c:c03::230 as permitted sender) client-ip=2a00:1450:400c:c03::230;
Message-ID: <54DCD3CF.1050300@gmail.com>
Date: Thu, 12 Feb 2015 17:24:47 +0100
From: Ilmen <ilmen.pokebip@gmail.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0
MIME-Version: 1.0
To: lojban@googlegroups.com
Subject: Re: [lojban] the myth of monoparsing
References: <CAO7bV+hZTwiquY0GcTA0xVjg2t+QnRt_DcY+XafDJYfa10L7ow@mail.gmail.com> <20150204124517.GA1243@kuebelreiter.informatik.Uni-Osnabrueck.DE> <CAO7bV+jCrZy4cO62r=jJXbdM67W9RZw=k5HBvML2ZgferZiJcQ@mail.gmail.com> <e157ff58-8ad1-4ca0-afc9-18cf7d2f352e@googlegroups.com> <CAO7bV+gV_p8w-de=N1DcfSt8CzkmSZ=bHp6H1AV3Fqm8wYTh0A@mail.gmail.com> <c04eb516-981e-4338-9728-a7e852523b4a@googlegroups.com> <CAO7bV+geB+=fq33PeF0duOhwFs-D93gYnyxTZo_SFK9rif-m=g@mail.gmail.com> <20150212123856.GA831@kuebelreiter.informatik.Uni-Osnabrueck.DE> <CAO7bV+j5dQogpWO9PZ-XkpNmAhWjX+gWhL0Juk0cQWJHZvQvcA@mail.gmail.com> <20150212150021.GB4686@kuebelreiter.informatik.Uni-Osnabrueck.DE> <CAO7bV+isGTfyufUbiZ8kRYQxSoNZivEeGtNn0b7ZEPJSB9nuBQ@mail.gmail.com>
In-Reply-To: <CAO7bV+isGTfyufUbiZ8kRYQxSoNZivEeGtNn0b7ZEPJSB9nuBQ@mail.gmail.com>
Content-Type: multipart/alternative;
 boundary="------------060804030908070300060502"
Reply-To: lojban@googlegroups.com
Precedence: list
Mailing-list: list lojban@googlegroups.com; contact lojban+owners@googlegroups.com
Sender: lojban@googlegroups.com
X-Spam_score: -1.9
X-Spam_score_int: -18
X-Spam_bar: -

This is a multi-part message in MIME format.
--------------060804030908070300060502
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable


On 12/02/2015 16:10, Gleki Arxokuna wrote:
>
> 2015-02-12 18:00 GMT+03:00 v4hn <me@v4hn.de <mailto:me@v4hn.de>>:
>
>     Everything you proposed in this thread to make lojban "polyparse"
>     is about semantic analysis, not about parsing.
>
>
> Adding polyparsing to Lojban is possible by removing some parts of its=20
> grammar and it has nothing to do with semantic analysis.
>

As for the "Fred saw a plane flying over Zurich" sentence, it's=20
certainly possible to have a monoparse if we allow more than two=20
branches per node, like the below:

sentence =E2=86=90 noun-clause finite-verb noun-clause verb-ing preposition=
=20
noun-clause
(Here the node "sentence" has six branches.)

But I guess that a true (and useful) parse analysis should not allow=20
more than two branches per node, so that for each word we know what=20
other word it attaches to. As the -ing verb-form can either represent an=20
adjectival present-participle (which attaches to a noun on its left) or=20
an adjectival present-participle (which attaches to the verb, I think),=20
it's not possible to produce a single **binary** parse tree.

Here's a sketchy binary+unary branching grammar:

sentence =E2=86=90 noun-phrase verb-clause
verb-clause =E2=86=90 verb-clause-2 adverbial-phrase
verb-clause-2 =E2=86=90 verb-phrase noun-phrase
verb-phrase =E2=86=90 verb-past
adverbial-phrase =E2=86=90 verb-ing preposition-clause
noun-phrase =E2=86=90 name | article noun | noun-phrase=20
adjectival-present-participle-clause
adjectival-present-participle-clause =E2=86=90=20
adjectival-present-participle-clause preposition-clause| verb-ing
preposition-clause =E2=86=90 preposition noun-phrase
...

If I made no mistake, the above grammar can parse the Zurich sentence as
=E2=80=A2 (Fred [{saw <a plane>} {flying <over Zurich>}])  -- adverbial pre=
sent=20
participle interpretation
or as
=E2=80=A2 (Fred [saw {<a plane> <flying (over Zurich)>}])  -- adjectival pr=
esent=20
participle interpretation

So I guess that the relevant concept here is "binary parsing", that is,=20
grouping the words so that each node contains no more than two child=20
nodes at once.

Just my two cents.

mi'e la .ilmen. mu'o


--=20
You received this message because you are subscribed to the Google Groups "=
lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an e=
mail to lojban+unsubscribe@googlegroups.com.
To post to this group, send email to lojban@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

--------------060804030908070300060502
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<html>
  <head>
    <meta content=3D"text/html; charset=3Dutf-8" http-equiv=3D"Content-Type=
">
  </head>
  <body text=3D"#000000" bgcolor=3D"#FFFFFF">
    <br>
    <div class=3D"moz-cite-prefix">On 12/02/2015 16:10, Gleki Arxokuna
      wrote:<br>
    </div>
    <blockquote
cite=3D"mid:CAO7bV+isGTfyufUbiZ8kRYQxSoNZivEeGtNn0b7ZEPJSB9nuBQ@mail.gmail.=
com"
      type=3D"cite">
      <div dir=3D"ltr"><br>
        <div class=3D"gmail_extra">2015-02-12 18:00 GMT+03:00 v4hn <span
            dir=3D"ltr">&lt;<a moz-do-not-send=3D"true"
              href=3D"mailto:me@v4hn.de" target=3D"_blank">me@v4hn.de</a>&g=
t;</span>:<br>
          <div class=3D"gmail_quote">
            <blockquote class=3D"gmail_quote" style=3D"margin:0 0 0
              .8ex;border-left:1px #ccc solid;padding-left:1ex">Everything
              you proposed in this thread to make lojban "polyparse"<br>
              is about semantic analysis, not about parsing.<br>
            </blockquote>
            <div><br>
            </div>
            <div>Adding polyparsing to Lojban is possible by removing
              some parts of its grammar and it has nothing to do with
              semantic analysis.</div>
            <div>=C2=A0<br>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    As for the "Fred saw a plane flying over Zurich" sentence, it's
    certainly possible to have a monoparse if we allow more than two
    branches per node, like the below:<br>
    <br>
    sentence =E2=86=90 noun-clause finite-verb noun-clause verb-ing preposi=
tion
    noun-clause<br>
    (Here the node "sentence" has six branches.)<br>
    <br>
    But I guess that a true (and useful) parse analysis should not allow
    more than two branches per node, so that for each word we know what
    other word it attaches to. As the -ing verb-form can either
    represent an adjectival present-participle (which attaches to a noun
    on its left) or an adjectival present-participle (which attaches to
    the verb, I think), it's not possible to produce a single <b>*binary*</=
b>
    parse tree.<br>
    <br>
    Here's a sketchy binary+unary branching grammar:<br>
    <br>
    <tt>sentence =E2=86=90 noun-phrase verb-clause</tt><tt><br>
    </tt><tt>verb-clause =E2=86=90 verb-clause-2 adverbial</tt><tt>-phrase<=
/tt><tt><br>
    </tt><tt>verb-clause-2 =E2=86=90 verb-phrase noun-phrase</tt><tt><br>
    </tt><tt>verb-phrase =E2=86=90 verb-past</tt><tt><br>
    </tt><tt>adverbial-phrase =E2=86=90 verb-ing preposition-clause</tt><tt=
><br>
    </tt><tt>noun-phrase =E2=86=90 name | article noun | noun-phrase
      adjectival-present-participle-clause</tt><tt><br>
    </tt><tt>adjectival-present-participle-clause =E2=86=90
      adjectival-present-participle-clause preposition-clause</tt><tt> |
      verb-ing</tt><tt><br>
    </tt><tt>preposition-clause =E2=86=90 preposition noun-phrase</tt><tt><=
br>
    </tt><tt>...</tt><br>
    <br>
    If I made no mistake, the above grammar can parse the Zurich
    sentence as<br>
    =E2=80=A2 (Fred [{saw &lt;a plane&gt;} {flying &lt;over Zurich&gt;}])=
=C2=A0 --=C2=A0
    adverbial present participle interpretation<br>
    or as<br>
    =E2=80=A2 (Fred [saw {&lt;a plane&gt; &lt;flying (over Zurich)&gt;}])=
=C2=A0 --=C2=A0
    adjectival present participle interpretation<br>
    <br>
    So I guess that the relevant concept here is "binary parsing", that
    is, grouping the words so that each node contains no more than two
    child nodes at once.<br>
    <br>
    Just my two cents.<br>
    <br>
    mi'e la .ilmen. mu'o<br>
    <br>
    <br>
  </body>
</html>

<p></p>

-- <br />
You received this message because you are subscribed to the Google Groups &=
quot;lojban&quot; group.<br />
To unsubscribe from this group and stop receiving emails from it, send an e=
mail to <a href=3D"mailto:lojban+unsubscribe@googlegroups.com">lojban+unsub=
scribe@googlegroups.com</a>.<br />
To post to this group, send email to <a href=3D"mailto:lojban@googlegroups.=
com">lojban@googlegroups.com</a>.<br />
Visit this group at <a href=3D"http://groups.google.com/group/lojban">http:=
//groups.google.com/group/lojban</a>.<br />
For more options, visit <a href=3D"https://groups.google.com/d/optout">http=
s://groups.google.com/d/optout</a>.<br />

--------------060804030908070300060502--