TLDR; how should I get in contact about doing a PhD with you?
First, you should definitely read the rest of this page before doing so!
Then, if you want to pitch a PhD proposal, the things you would put together to send me would be:
- a concise research outline of what you propose to work on and why (typically 2–3 pages, more detail below)
- an up-to-date CV - it's most useful if that covers your academic training, including courses and grades (esp. linguistics, programming, speech technology, machine learning etc.), and other relevant experience, which might include, for example
- jobs
- projects
- links to any publications
- an indication of your proposed funding route(s), entry date and expected home/overseas student status.
So you want to do a PhD?
Thanks for your interest in the possibility of studying for a PhD with me here at the University of Edinburgh!
Deciding to apply for a PhD is a big deal. Developing a competitive application takes significant effort (from both of us!). Working on a PhD if your application is successful likewise takes a big investment over several years, again for both candidate and supervisors.
There are multiple important factors which must align for a PhD place to be successful, including: relevance, feasibility and promise of the proposed topic, fit with the supervisor(s) and wider lab, a competitive academic/experience profile, and funding.
I'll mainly say things about both research topic and funding here. My aim is to give helpful guidance for next steps if you'd like to "pitch" a PhD proposal. It's a bit on the long side, but bear with me!
What research topics will you supervise?
The range of topics I'm interested in supervising is quite broad. In short, the main speech technology topics I have been most involved with so far include:
- anything related to speech generation, especially in low-resource settings (e.g. lightweight models, and/or minoritised or under-resourced languages);
- speech evaluation (objective and subjective);
- accent modelling;
- pronunciation modelling;
- articulatory data and modelling;
- child speech therapy and other medical applications of speech technology.
More detail of my work in those areas can be found from looking at my publications and homepage.
But don't think of that as an exhaustive, definitive list - I'm open to considering proposals for things I haven't considered before! If you can give a clear and concise description of what you want to work on and why, I can look at it and get back to you whether it's something I can work together with you on or not. That is where the Research Proposal document comes in.
Importance of a research proposal
PhD funding sources generally require you to write a Research Proposal as part of an application. This is a short document which covers several things, for example it:
- outlines the research topic
- points to prior work in the literature and gaps in the current state of the art
- presents other motivation for why you feel it is important to work in the area
- gives an overview of the approach(es) you propose to follow
- sets out a rough expected timeline for the research work
- lists expected risks and actions that can be taken to mitigate them
- may address preliminary considerations of data access, ethics, and suitable collaboration (esp. for projects involving clinical populations, children, personal speech data, or work with language communities)
Though not all PhD funding sources require exactly that, some document like that is the best starting point for any conversation between us when someone proposes to study for a PhD here with me as supervisor.
Think of it as a "project pitch" for research you want to work on together. Once you've set out your initial ideas - what you want to work on, why and how - that gives a foundation for discussion we can build upon. A lot of the point of a PhD is independence and initiative to tackle research challenges, so it's important for that initial impetus to come from the PhD candidate. If what you want to do seems compatible with my own expertise and interests, I'll be very glad to discuss it to develop the proposed research further. On the other hand, if what you want to work on doesn't seem a strong match, then I'd try and point to others that might be a better fit. Either way, I try and give constructive feedback!
The format of the document is up to you. It should be concise, though it'd be hard to make it less than 2-3 pages without leaving out important information. It would usually include references to other work, and a skeleton work plan.
Funding
In terms of funding, unless I am advertising a PhD place (in which case I'll already have funding secured, and it will be made clear in the advert), it will be necessary to secure funding. There are a few options I can point to, discussed in turn below.
University of Edinburgh funding
I am affiliated with two institutes here at the University of Edinburgh: The School of Informatics, and the school of Psychology, Philosophy and Language Science (PPLS). Both of these have a number of studentships available for applications each year to support ~3 years of PhD study:
- Within Informatics, I am affliated with the Institute of Language, Cognition and Computation (ILCC) and so can supervise students eligible to apply to the ILCC PhD programme.
- In addition, as a member of the Linguistics department in PPLS, I can supervise students applying to the PPLS PhD programme.
These funded place are highly competitive, particularly for international applicants. Two of the PhD students I've worked with previously have been successful applying to the above department-linked programmes and funding, though. All the rest have been students on a CDT, more on that next...
Centre for Doctoral Training (CDT) programmes
These are multi-year, cohort-oriented PhD training programmes funded by the UKRI funding agency. They are ~4 year programmes, which includes time to take courses equivalent to a 1-year MSc degree. In some CDTs, the programme starts with a year of courses, then 3 years for research, while in others the courses can be spread over multiple years, called an "Integrated Programme". Each year around 10-15 students are selected to go through the 4-year programme together as a cohort. We have ~5 of these CDTs running here currently (which typically each have 5 intake years), which I've listed below ranked in descending order of the relevance to the research topics I'm usually involved with:
- Developing Responsible NLP (DR-NLP)
- Machine Learning Systems
- AI for Robotics
- Sensing, Processing and AI for Defence and Security
- Biomedical Innovation
You should look at the remit for each, and also your eligibility, and decide which is/are the best fit for the research agenda you want to pursue. As with the Edinburgh University studentships, competition for these CDTs is strong (and especially so for international students).
External sources of funding
These including from companies, or national bodies in an applicant's own home country, for example. It is definitely worthwhile looking for these, and if your proposed research topic is right and I'm positive we could work well together, I would of course help however I can in that application process.
As a final comment on funding, I would strongly discourage self-funding a PhD unless you are very secure financially, have funding for the full duration of the programme ready and have carefully considered the financial implications - because money worries while trying to do a PhD would be just too much of an uphill battle. Please also note, you might already have been successful in securing funding but that does not mean I will automatically agree to sign up as PhD supervisor. As for all academics here, the research topic, and a positive sense that we can work together successfully, is of paramount importance when deciding whether to support a PhD application.
Application timing
In principle, it is possible to start a PhD at any point throughout the year. But PhD funding streams usually have application deadlines in the Autumn/Winter of one year for a start in the September of the following year. (Do check specific deadlines and eligibilty criteria for funding schemes carefully!)
Capacity
I should also note that my capacity to take on new PhD students naturally varies from year to year, and even if I'm very interested, it doesn't guarantee that I will be able to supervise a PhD starting in a particular year unfortunately.
Still want to do a PhD?
Then I'll be glad to hear from you! Look back at the top of this page and see what to include in an initial contact - thanks!
.