Getting started with the discussion section


Last year I wrote a post called How do I write the discussion section? This post was in response to an analysis of blog traffic which showed that 75% of the writing related searches that led people to the Thesis Whisperer wanted to know about this section of the thesis. The discussion section is the ‘problem child’ of the thesis because it asks you to be creative. In How do I write the discussion section? I covered what could go in the discussion section and in what order, but I did not address the creativity problem.

People have spent lifetimes studying creativity. There are reams of books on the subject. However, as yet, no one has invented a magic creativity elixir.

Which is a bit of a shit really.

So – how can you become a more creative researcher? I think the secret lies in thinking about creativity as a process, not an accident, inspiration or something that just ‘happens’. Becoming creative on demand is possible and a key skill for all researchers to cultivate.

One of my favourite books on creativity is Daily Rituals: how great minds make time, find inspiration. In it, Mason Curry shares the daily schedules of famous writers, musicians, engineers, architects and philosophers. Each person’s creative routine is explained in just a few pages. Reading all these snippets together is a fascinating insight into how different people, in different places, and with different pressures, carried on with the work of being creative.

Some people were more free to explore their creativity than others. Jane Austen, like many women of her time and culture, was not expected to have an occupation. She did not have a dedicated place to write or think and had to do it in the family spaces (many PhD parents can relate to I’m sure). According to her nephew, Austen wrote on “… small sheets of paper which could be easily put away, or covered with a piece of blotting paper”. She  was protective of her work in progress and deliberately sat by a creaking door so that she would have early warning of anyone coming, enabling her to put away her writing and pick up her sewing.

Austen is a decided contrast to Patricia Highsmith, writing some 120 years later. Highsmith  wrote great thrillers, including Strangers on a Train and The Talented Mr Ripley. What I like about Highsmith is it seems she was simultaneously totally organised (she always wrote for 3 – 4 hours a day and could get a good 2000 words down in that time) and a complete slob, as her biographer Andrew Wilson recounts:
Her favourite technique to ease herself into the right frame of mind for work was to sit on her bed surrounded by cigarettes, ashtray, matches, a mug of coffee, a doughnut and an accompanying saucer of sugar. She had to avoid any sense of discipline and make the act of writing as pleasurable as possible. Her position, she noted, would be almost foetal and, indeed, her intention was to create, she said, “a womb of her own.”
(God, how I love her. I want to be her.)

I trained as an architect and was taught not to wait for creativity to come knocking at your door. One thing I learned in architecture school is that when you are stuck, it can help to switch tools. Stop trying to draw the building: make a 3D model in cardboard. If that doesn’t work, make the model in clay or plastic instead. Draw in pen instead of pencil. Stop drawing with a ruler and go freehand. Colour it in with crayons; wash the drawing over with gauche.  Make a collage. Drag your plan around as you photocopy it. The options are endless: half of the fun of being at architecture school was dreaming up new creative processes.

I left the profession some 20 years ago for lots of reasons, and don’t miss it, but the down to earth attitude to creativity stayed with me. The reality of work in an architects’ practise runs somewhat counter to what you see in the movies, where architects are often portrayed as tortured geniuses. Architects don’t have time to romanticise the creative process: they just roll their sleeves up and get on with it.  In real life, they just grab a pen and try to get the building done by lunchtime.

Once you learn to be creative on demand, you lose your fear of not having enough ideas. Generating ideas for buildings becomes almost effortless but you are not out of the woods because most edesign ideas are crap.

In architecture school you are taught to assume most of what you make is shitty. This teaches you to hold on to ideas lightly and not invest a lot of effort until you are sure they are not crap. This is why architecture teachers are obsessed with ‘process’. If you bring a set of plans complete to the toilet tile layout to class, you will be sent back to the drawing board, literally. Design teachers want to see messy sketches and lots of ‘iterations’; showing you have thought through a lot of alternatives before settling on your final design scheme. You only get to finalising the toilet details after months of work testing the design idea out. They call this process ‘working through the ideas’.

The sketch books that architects sometimes publish are usually carefully composed fictions – most of the ‘working through the ideas’ drawings are very ugly and end up in the trash. I’ve always approached the creative act of writing up my research results in the way I was taught to design buildings and found it works well. By anyone’s measure I am a productive person. I’ve written, co-written or edited seven books in the last six years as well as running a weekly blog for 10 years. And yet I throw out at least half of my writing – maybe more.

This way of working, where you try to have lots of ideas, but hold on to them lightly, is very freeing. My best advice for the discussion section is not to agonise about what to write. Commit every idea – especially the stupid ones – to ‘paper’, test them out first, and then decide what to throw out. When I say ‘commit it to paper’ I don’t mean write it down formally, in polished prose, but in quick, unfinished ‘chunks’ – a term I stole off Pat Thomson and Barbara Kamlers’ book ‘Helping Doctoral Students write’ where they recommend ‘chunks and not chapters.


View this post on Instagram


A post shared by Thesiswhisperer (@thesiswhisperer)

Just like an architect who picks up a different pen, or mucks around on the photocopier, you can ‘switch tools’ and use diagrams or exercises to test out your half-baked ideas quickly. My colleague Victoria Firth-Smith has a good writing exercise where she asks people to assign a genre for their thesis and write it on a post it note. This is a great way to find the story line that runs through your discussion section.

You can find a genre for your thesis by imagining you are in a bookstore: what section would a reader find your thesis? Is it: crime, romance, literary fiction, self help, history, crime? Or you can think about your thesis as a narrative type: is it a tragedy, a comedy, a murder mystery or a manifesto? You can have a look at the list of TV tropes for more inspiration.

The next step is to write a little story using this narrative form to explain your research to someone else, maybe at a dinner party. Who are the main characters in your story? What are the main plot points? Who are you as a story teller?

Here’s an example, using my own research. Our PostAc app uses machine learning to find industry jobs for PhD graduates. During this research, we discovered that 80% of employers who are looking for a person with advanced research skills did not use the term ‘PhD’ in their ad. Employer are looking for someone like a PhD graduate, but without realising it. Why don’t they put ‘PhD’ in the ad? All kinds of reasons as it turns out – some of which I have documented in a previous post here where I talk about an ‘anti-PhD’ attitude.

Lots of people wrote to me after I published this post, reporting that employers kept telling them they were ‘over qualified’ or ‘not a good cultural fit. This suggested another avenue of research: are employer attitudes a barrier, or is there something in the way PhD students present themselves that puts employers off? Or both? Here’s my research story in different genre styles:

Narrative type: Police procedural

PhD graduates report feeling traumatised after being interviewed by non academic employers. After these interviews, PhD graduates feel like they have no place in the non-academic workforce. We, the detectives, are trying to find out if a crime has actually occurred. We do this by talking to people on both sides of the conflict. Exactly what kinds of trauma are PhD graduates experiencing? What type of crime do they think the employer has committed? Do the employers even think they have committed a crime? If not, how do they explain their behaviour?

Narrative type: Doing in the Wizard (where what looks like a supernatural phenomenon has a mundane explanation)

Employers keep ignoring the the brilliance of PhD graduates because they don’t understand who and what they are. In reality, employers are quite aware of some of the downsides of PhD training and are conciously  choosing employees with experience of the workplace context. This is because what we call ‘transferable skills’ are not so transferable after all.

For me, this is a very clarifying exercise. By switching into a narrative mode I was able to think through and articulate a couple of insights about my data which I didn’t realise were there. I know it sounds weird, but try it and see if it works for you.

The discussion section encourages us to lean in to the creative part of our work – something we are often not explicitly taught. If the couple of techniques I’ve suggested here don’t work for you, spend a bit of time asking your supervisors and others in your department how they write the discussion section. In my experience, academics have all kinds of tricks and hacks for getting to the research insights. I suspect that, just like with architects, people think these tricks are silly or ugly and don’t think to share them unless you ask. Give it a go!

I’ll see you in a month, but in the meantime, if you want more Whisperer, check out my podcast with Dr Jason Downs. You can find all the subscription options on our On the Reg page.