WEBVTT

00:00:00.000 --> 00:00:05.220 align:middle line:90%


00:00:05.220 --> 00:00:08.760 align:middle line:84%
My research is on literature
that was written and published

00:00:08.760 --> 00:00:11.160 align:middle line:90%
between 1911 and 1928.

00:00:11.160 --> 00:00:12.810 align:middle line:84%
And I'm using public
domain literature

00:00:12.810 --> 00:00:14.640 align:middle line:84%
so that I don't
have any copyright

00:00:14.640 --> 00:00:17.810 align:middle line:90%
issues with publishing the PhD.

00:00:17.810 --> 00:00:19.830 align:middle line:84%
And as far as
manipulating that data,

00:00:19.830 --> 00:00:22.590 align:middle line:84%
I use computational
linguistic software

00:00:22.590 --> 00:00:26.310 align:middle line:84%
from Lancaster University, and
a few other local universities,

00:00:26.310 --> 00:00:29.380 align:middle line:84%
to analyse these stylistic
tendencies of these authors.

00:00:29.380 --> 00:00:35.130 align:middle line:90%


00:00:35.130 --> 00:00:37.470 align:middle line:84%
It is important, I think,
to make sure your data is

00:00:37.470 --> 00:00:38.700 align:middle line:90%
valid in your research.

00:00:38.700 --> 00:00:42.150 align:middle line:84%
And in my case, because I
was examining 95 novels,

00:00:42.150 --> 00:00:45.690 align:middle line:84%
I used software called
Multi-dimensional Analysis

00:00:45.690 --> 00:00:47.610 align:middle line:84%
Tagger, and that
software identifies

00:00:47.610 --> 00:00:50.920 align:middle line:90%
the genre of the text.

00:00:50.920 --> 00:00:53.490 align:middle line:84%
And what I found was of the
95 novels, eight of them

00:00:53.490 --> 00:00:55.440 align:middle line:90%
were tagged involve persuasion.

00:00:55.440 --> 00:00:57.270 align:middle line:90%
They were not tagged as fiction.

00:00:57.270 --> 00:00:58.750 align:middle line:84%
So then I had to
make the decision

00:00:58.750 --> 00:01:01.320 align:middle line:84%
of do I include them in my
corpus, and in my research,

00:01:01.320 --> 00:01:02.550 align:middle line:90%
or do I exclude them?

00:01:02.550 --> 00:01:07.780 align:middle line:90%


00:01:07.780 --> 00:01:10.300 align:middle line:84%
I do think there are different
ways of interpreting findings

00:01:10.300 --> 00:01:12.610 align:middle line:84%
in my field, which
is corpus stylistics.

00:01:12.610 --> 00:01:15.850 align:middle line:84%
The field has only been around
for a couple of decades,

00:01:15.850 --> 00:01:18.910 align:middle line:84%
and all I think research should
be grounded in scholarship,

00:01:18.910 --> 00:01:21.130 align:middle line:90%
and also originality.

00:01:21.130 --> 00:01:25.120 align:middle line:84%
So I do find myself sometimes
not knowing which way to go,

00:01:25.120 --> 00:01:27.750 align:middle line:84%
whether something is valid,
or should be discarded.

00:01:27.750 --> 00:01:31.330 align:middle line:84%
And it's at those times where
I tend to seek outside advice.

00:01:31.330 --> 00:01:33.850 align:middle line:84%
Sometimes from the
supervisors, but also sometimes

00:01:33.850 --> 00:01:35.980 align:middle line:84%
from outside the
supervisory team.

00:01:35.980 --> 00:01:41.420 align:middle line:90%


00:01:41.420 --> 00:01:45.080 align:middle line:84%
I don't really try to
avoid bias in my research.

00:01:45.080 --> 00:01:49.400 align:middle line:84%
Because again, it's a new field,
corpus stylistics, hopefully

00:01:49.400 --> 00:01:51.710 align:middle line:84%
the research is
grounded in scholarship.

00:01:51.710 --> 00:01:54.050 align:middle line:84%
And I think it's OK to
have unexpected results.

00:01:54.050 --> 00:01:57.830 align:middle line:84%
And I think sometimes it's good
to have unexpected results.

00:01:57.830 --> 00:02:00.650 align:middle line:84%
As an example I was
doing a presentation

00:02:00.650 --> 00:02:03.650 align:middle line:84%
in a paper on Americanism
and literature

00:02:03.650 --> 00:02:09.280 align:middle line:84%
and how in the era from 1911
to 1928 America was portrayed.

00:02:09.280 --> 00:02:14.270 align:middle line:84%
And in that research, I found
some very unusual statistics

00:02:14.270 --> 00:02:17.300 align:middle line:84%
using computational
linguistics software

00:02:17.300 --> 00:02:19.670 align:middle line:84%
that's just absolutely
phenomenal, and somewhat

00:02:19.670 --> 00:02:20.413 align:middle line:90%
groundbreaking.

00:02:20.413 --> 00:02:21.830 align:middle line:84%
So at first it was
like this isn't

00:02:21.830 --> 00:02:23.630 align:middle line:90%
the way I wanted things to go.

00:02:23.630 --> 00:02:25.200 align:middle line:84%
And then it's
like, wait a second

00:02:25.200 --> 00:02:28.030 align:middle line:90%
this is really interesting.