"Real Artists Ship"

Colin Johnson’s blog


Archive for the ‘UniKentComp’ Category

Immanence in Improvisation

Tuesday, December 25th, 2012

Went to an interesting talk by George Tzanetakis earlier in the week at the DMRN meeting at Queen Mary. He was discussing how systems such as the Kinect could be used to extend the performance of an acoustic instrument by adding gesture recognition to control the electronic post-processing of a sound. Also saw a performance by Imogen Heap a few weeks ago along similar lines.

This got me thinking about how we explore sound in improvisation. I do a lot of “free improvisation” using the bassoon, and an interesting aspect of this is how I explore transformations of the current sound whilst playing, without actually making sound. One aspect of this is what we might term immanence, that is, the feeling of a new sound “on the lips” before it is actually made. My approach to free improv is primarily textural, finding musical textures that fit alongside other improvisers in the group, which provide a new direction for the music, or which set out a radically new direction for the developing improvisation. By moving a key on the instrument, or adjusting pressure on the reed, I can start to feel when a sound is about to “break” into another sound, and get some sense of what that sound is likely to be—whether it is going to be a rougher sound, or whether it is about to break out into a pure, high harmonic, or whatever.

This sense of immanence is largely absent from interfaces for electronic instruments. Whilst many kinds of playing surfaces and unusual interfaces exist, they offer little back to the player in terms of pre-aural feedback about what sound-quality they are likely to move into if they move in a particular direction in the sound-space of the system generating the sound. Creating such interfaces, and thinking about how to provide such immanence, would make an interesting research project.

Viva (1)

Sunday, December 16th, 2012

An overseas colleague who is examining a PhD in the UK for the first time asked me for some advice on the viva. I thought that this would be worth sharing—some is specific to the Kent system, but generally it is generic. Enjoy.

You need to fill out a preliminary report form before the viva.

On the day of the viva we will meet and briefly discuss the structure of the viva and our general impressions before we ask the candidate in. The viva sometimes begins with a short overview presentation by the candidate. I am generally in favour of this, as it puts the candidate at ease.

There is then a decent length of time (usually around two-three hours) of questions from the external and internal examiners, often starting with some general questions, then drilling down into very specific detail (very specific questions along the lines of “on page XX, you said…can you explain why…” are usual). Typically this is structured by going chapter-by-chapter through the thesis. The question session often ends with a couple of general questions and an opportunity for the candidate to tell us about anything that they feel we have missed.

It is usually just the three people (internal and external examiners and candidate) in a room. If the candidate and examiners want it, the supervisor can attend, but I would advise against this, it seems offputting to the candidate. There is also an option for an “independent observer” to be present which wouldn’t normally be used unless there are no internal examiners.

There is, despite what you might have seen in films set in Oxford in the 1930s, no need to dress up like Henry VIII or carry a ceremonial sword. The candidate will probably wear a tie which is shocking after seeing them in T-shirt and jeans for the last four years.

At the end of the questioning the candidate is sent to pace up and down anxiously out the room or go out of the building and smoke an unfeasible number of cigarettes whilst the internal and external examiner make a decision, there-and-then. The nine different options that we have are listed in Part B of the aforementioned form. A brief guide to these options:

Option 1: Pass with no corrections. Very rare. Examiners like to ask for at least a few corrections to show that they have read the thesis.

Options 2 and 3: Minor corrections of two different lengths. Probably the most common outcome. Both of these basically indicate to the candidate that the thesis is basically fine. The first one is for minor points (e.g. grammar/spelling, redrawing graphs and diagrams, putting in some references, adding a few explanatory sentences here and there; the second option is if the corrections are more substantial but the basic argument of the thesis is clear and sound.

Option 4: Resubmission after a year. A not-uncommon option, means that the examiners are as yet unconvinced that there is a PhD in this work but that there is a reasonable chance that something could be gotten out of it with a decent amount of work. This is the worst result that should be given the first time round – we should, unless the thesis is a complete travesty, always give the candidate a second chance.

Options 5a and 5b: retake oral exam and/or take a written exam. To the best of my knowledge, these have never been used in the history of the universe/university.

Options 6, 7, 8. Various levels of failure or fallback awards. Shouldn’t be chosen this time round but would be options for a candidate who has been offered option 4 and has failed to get the thesis up to the level required.

We then fill out this form and call the candidate and supervisor back into the room and tell them the result. There is a detailed report to be filled out then, which we should ideally do there-and-then or could do by email a little later if you have to dash away. This should ideally set out precisely what we expect in the way of corrections (obviously, in the case of a resubmission where the original thesis was weak, this might not be so easy).

Getting Meso-scale Information from People

Thursday, November 8th, 2012

The National Student Survey (NSS) has been running for a number of years now, and, in common with similar exercises carried out by many organisations, uses two methods for getting feedback from students about their experience at university. The first of these are Likert scales asking for a numerical rating on a number of questions. For example, students might be asked to rate the quality of the library service on a scale from 1-5. These numbers are then averaged, at a course or institutional level, and these averages compared to other courses or institutions in the sector. The second is to ask for free text comments, e.g. positive experiences and suggestions for improvements.

These capture two kinds of information, each with disadvantages. The numerical questions capture macro-scale data about where problems are; but, the information is not at sufficient detail to inform action. Merely knowing that, say, the library is weak doesn’t inform us about what the concerns are: is this a problem with the availability of books, the opening hours, the quality of the space, or what? However, the free-text comments measure things at a micro-scale. Each comment represents the individual viewpoint of an individual student, which might be a shared concern of the whole cohort or might be a personal, unshared concern.

What we need for action is meso-scale information, information about the broad-brush details of what a large proportion of the student body are interested in. One way that we have found useful in getting this sort of meso-scale information is to use the results from the NSS in focus groups of students. By discussing the results with students, we can ask questions about why they think a particular low or high score on a numerical question got that way, and what might be done to improve it. Similarly, with the micro-level free text questions, we can ask whether a particular response is a common concern of a number of students, or which free-text responses stand out as being general concerns. Using the NSS results to provoke discussion within such a focus group seems to give a better quality of discussion to these points; it gives a starting point, a “focus” if you like, to the group’s discussion.

On Newsletters, Shoddy and Fine

Thursday, October 25th, 2012

Lots of sub-units of large organizations try to create newsletters. I am thinking of things at the scale of a university department or faculty. My experience with these is that they start out with a great fanfare, and the first two-three issues, produced as glossy printouts or pdf files, look fantastic and are full of interesting articles. However, after that, production tails off; people have already disburdened themselves of things that they want to write, and the overworked administrator who has taken on the job of producing the newsletter has other important things to do.

One problem is that these newsletters are fairly unfocused in their audience; are they for the consumption of that sub-unit internally, or are they meant to be for a wider audience. This is often unclear. In many cases, a news stream or blog (perhaps different ones for different audiences) might be better; but, again, it still needs to be used and to have some sense of audience.

Something that worked very well, though, was what I think of as the “shoddy newsletter” for internal communication. When I was a Head of School, I would keep a text file at hand, and update it with ideas that I was thinking about, things that I wanted feedback on, and snippets from emails that I had received from elsewhere in the University or outside. This would then get emailed to staff as an “update” every week. I found this effective; basically, people were kept in the loop without being burdened with lots of emails being sent out. If I were to do this again, I’d probably do one for students, too.