Open Panel on Non-traditional Authorship Attribution Studies

Joseph Rudman
Department of English
Carnegie Mellon University
Pittsburgh, PA 15213
Rudman@cmphys.phys.cmu.edu

During the last three or four ACH-ALLC conferences, many of the people working in the areas of non-traditional authorship attribution, stylistics, and other related fields met over coffee, discussed various problems, and expressed the need for a more formal gathering. This group included the following (and I apologize if I missed anyone -- no notes were taken):

Harald Baayen, Richard Forsyth, Bernie Frischer, Lyman W. Gurney, Penelope J. Gurney, David Holmes, Thomas B. Horton, Joseph Rudman, and Fiona Tweedy.

I have been asked by many of these practitioners to arrange a panel at this meeting. It is an open panel -- anyone may request an item be placed on the agenda and lead the discussion for that item. The final "members" of the panel will include many practitioners not on the above list.

There will be no papers presented during this session. It will be a moderated, modified open discussion on the state of non-traditional authorship studies, the direction these studies should be going, and how to get there.

Practitioners were and invited to submit short (~5 minutes) summaries of leading edge work in progress for possible inclusion in the session. No more than four of these will be scheduled for presentation in the session. Everyone was invited to post "work in progress" summaries by May 30, so that these would be digested before the session convenes.

Talking points were made available before the meeting to all interested members of the ACH-ALLC. We did not want to dictate which talking points should be discussed. We have solicited suggestions for talking points, questions, and/or comments from all interested parties and only then was an agenda set.

Since the final agenda was not be set until shortly before the meeting, we can only give a conceptual overview in this abstract.

The following were "seed" talking points and were made available to all interested parties:

  1. Are any published non-traditional authorship attribution studies valid?

  2. What elements must a non-traditional authorship attribution study have to be valid?

  3. How can we answer Furbank and Owens' statement, "...stylometry cannot be classed as a `science,' and must presumably be thought of only as an aspiration."?

  4. Why can practitioners not agree on which statistical tests should be used?

  5. Is quantifiable style a closed system?

  6. How do we account for the changing concept of "author?"

  7. What is wrong with non-traditional authorship studies?

  8. What is right with non-traditional studies?

  9. What, if anything, should be done by this group to bring some kind of order to the field?
One topic that was suggested by Professor Gerald R. McMenamin and put on the agenda after the conference proposal was accepted and before the deadline for this abstract is:
The Value of Forensic Linguistics to Non-Traditional Authorship Studies

This entire panel session will be recorded, transcribed, and disseminated. An edited and expanded version of the transcription is planned.

We will discuss if this should become an annual session (subject each year to the approval of the appropriate committee) during the ACH-ALLC conference.

We do invite any interested conference attendee to join in this open panel discussion.