
Jonathan and Sam here, with just a quick note before Interview. Number. One!
We’ve been known to talk about the fact that we did our PhDs in the same discipline, in the same cohort, at the same University. And how we then ended up working in the same organisation at one stage. And how we've found this really fascinating, considering our quite different career journeys from that same PhD starting point. Well, we’re now about to double down on that…
When we put the call out for people with PhDs willing to share their career stories for this newsletter, it was lovely to see the familiar names and faces of some of our PhD alumni in our LinkedIn mailboxes.
These people with PhDs from our discipline, our cohort, our Uni. There are some parallels you could draw in the roles we’ve ended up in too, but we think it’s the differences that are really interesting. And so, we’re kicking off the interviews in this series with a couple of our former PhD colleagues. The interviews were a bit of top-quality, reunion-style fun too!
First up, we introduce Dr Stephen Pritchard. Stephen has plenty of wisdom on the topic of career switching, having already managed a double switch: from an early career in renewable energy policy, into a PhD in Cognitive Science focusing heavily on computational modelling in reading acquisition, to now being back in a role with a renewable energy focus at Evergen but with the skills from the PhD adding a whole new dimension to his practice.
In this interview we dig into this double switch. We talk about pragmatics and trade-offs when an academic life collides with all that other life stuff. And we hear more about Stephen's goal to unravel the mysteries of the mind.
Jonathan McGuire: I'm going to start by focusing on the time before your PhD, because you had a bit of a career before you jumped into academia. What were you doing before your PhD and what inspired you to pursue a doctorate?
Stephen Pritchard: The main thing I studied in my undergrad was electrical engineering, but I have an Arts/Electrical Engineering combined degree with sociology as my major and a bit of linguistics. From this, I got into renewable energy back in the late 90s, early noughties. Renewable energy was exciting, but there weren't a lot of jobs. I ended up working for the State government in policy-related work, initially supporting renewable energy then moving into more general climate change work: a bit of energy efficiency, a bit of solar power, a bit of wind power. But I still felt a bit unsettled – there was something missing, but I wasn’t quite sure what that was. After a few years, I was doing some computational modeling for my work, creating these really detailed models of household energy consumption, where you could adjust things like occupancy and kinds of appliances, and it would spit out how energy was being used in the house. And I realised that I really liked computational modeling.
I’d always been really interested in questions like ‘what is the mind?’, ‘what is cognition?’, ‘what is artificial intelligence?’ and I’d been reading some Oliver Sacks and Douglas Hofstadter among other things. Then suddenly all the pieces felt like they had come together. I didn’t know what cognitive science was up until this point, but it was a revelation for me that psychology isn't just clinical psychology; there's research psychology, and there are theories that try and treat the mind as a kind of machine and theorise how it works. I felt really clear that this is what I wanted to be doing, and with a wave of enthusiasm behind me, I managed to talk myself out of my background in electrical engineering, renewable energy and government and talk myself into a PhD in cognitive science.
I took the confidence that comes with thinking that you found what you want to do, and a lot of luck, and I found a PhD supervisor who just happened to want someone with a computational modeling background. And they decided that teaching me cognitive science would be easier than teaching a psychologist computational modeling. Certainly the Open University courses I did as pre-career-change preparation, and my undergrad linguistics helped too, but I think it was quite fortunate to be able to talk my way across disciplines like that instead of going back and doing another undergrad. In fact, I remember when I first contacted this PhD supervisor with my plan, and it was a huge deal for me to have even sent that email, and he replied within 15 minutes and said, 'I think you'll have to go back into an undergrad'. It was fortunate that I was old enough, bold enough and determined enough to reply, 'No, I don't want to do that. I think I've done enough study and that this should be appropriate for this, this and this reason' and he said 'Oh, I didn't think about it that way. Okay, come and talk to me.'
JM: We did cognitive science PhDs in the same cohort at the same uni, but I think it's kind of wonderful that you came from this computational modeling background, and I came from a social sciences background and CogSci turned out to be this quite broad church that could encompass both. That’s how we got to know each other and something that I think is really valuable about the field. So, once you got into the PhD program, what exactly were you researching? And can you tell us a bit about what you found?
SP: My topic was computational modeling of the cognitive processes employed when we learn how to read and when we read. My PhD supervisor had developed an existing computational model of skilled reading, which explored computationally a theory that there are two mental mechanisms broadly involved in reading: a sort of whole word processing mechanism and a sub-lexical mechanism that relies on letter sound correspondences. My particular little niche was that I augmented that model so that it could also say something about how we acquire reading. There’s a particular theory of acquiring reading skill called the ‘self-teaching hypothesis’: it’s the idea that we learn more words than could possibly ever be individually, directly taught. So, there must be some mechanism to self-teach. And my computational model was an implementation of how that might work.
I explored several computational model paradigms for my PhD, and by the end had built and tested a new computational model of how sub-lexical knowledge and context supports self-teaching of new written word knowledge. Along the way, I also even did some experimental work to support my computational modelling. This experimental work involved building a dataset of how skilled human readers pronounce pseudo-words like SKONK or BREIK. A big part of cognitive computational modelling is assessing your model against human behaviour, and identifying the possible insights and possible shortcomings of the theory tacitly suggested by the model architecture. This is down in the detail I guess, but I found it particularly fascinating how both humans and the model we built learned “weird” words, like heterographic homophones (different spellings, same pronunciation, like BREAK and BRAKE), homographic heterophones (same spelling, different pronunciation, like WIND) and a potentiophone (an irregularly pronounced word that, if you try and pronounce it according to typical letter-sound correspondences, sounds like something that is different, but also still a word! (like if you pronounce CORPS the way it looks, you get the real spoken word “corpse”.)
JM: And, because a PhD like that is not intense enough, during your PhD you got married, had a couple of kids; I'm not sure if you were working as well? Just how did you juggle all of that?
SP: Ah yes, good question. When I started my PhD, I had no children and I was unmarried. I had a girlfriend, and we were really supportive of each other’s dreams. I got a PhD scholarship, and that was my income. I wasn’t working besides that. And during my PhD, we did get married and had a son and then our second child, a daughter, we had just before I graduated.
Somehow it wasn’t really as hard as I thought: because my wife was working, we were ok financially, we had to belt-tighten a lot compared to when we were both working, of course, but I think we were the right kind of people to be able to cope with that. And I could use the flexibility that you get with studying to good effect with parenting. Not saying that I did most of it, my wife did most of it, we were fortunate that she was able to take a long stretch of parental leave. But I was better able to balance my PhD with having lots of time with my son and being involved in swimming classes, play, nighttime soothing, and what was happening at his day care and all those things. When I started, I didn't intend to take a long time, I thought 'I'm old. I can't afford to take a long time doing a PhD. I'll do it really quick.' and I took longer than I thought, which was a bit over four years. But I think I did okay, given I had a newborn at that time.
JM: The closest I got to responsibility while I was doing my PhD was keeping air in my bicycle tires. So I really commend you for managing to pull that off.
SP: Well when the scholarship runs out after three and a half years and you've got a child, a family, I just couldn't afford to dawdle. In a way that probably helped. If you need extra motivation to actually get through those difficult last six months of actually writing the thing, that helps. It was really hard for my wife, though, she sacrificed a lot for me to do that. Doing a PhD represents a big financial sacrifice, so there’s that which impacts both of us, but a PhD is also very much a mental battle, lots of emotional ups and downs and my wife was exposed to and supported me in that, while caring for a newborn! I have developed good habits of taking personal responsibility for my own mental health, and also took advantage of the counselling services that are available to university students – something I would recommend to anyone and especially PhD students. It’s self-care, it is available, it helps.
JM: And then you stuck around in academia for a while. You did a couple of postdocs; I know you kept publishing for a bit. What was that like?
SP: That was actually a lot harder than my PhD in many ways. I went into my PhD with a mixture of confidence and naivety, I think, believing I could make this big career change and feeling like 'now that I've got my passion, nothing's going to stop me'. And it did start seeming like naivety during my postdoc years. Married, with a couple of kids, financial concerns as we were trying to buy a house, and then realising that passion is not enough. Instead, you have to be – one – really, really good because it's so competitive and – two – have a bit of good fortune as well. I think that luck probably does play a part in some postdoc trajectories, some picking topics that just generate loads and loads of interest, some managing to hit upon a really good job. And with young kids and aging and infirm parents we didn't want to move interstate or internationally, and not moving really limits your postdoc options as well.
In the early years I had a couple of postdocs that were fairly comfortable. And then it became a lot harder to find things that I wanted to do, and to stay employed full time. I didn't go into cognitive science specifically to research reading, I went into cognitive science to use computational modeling to unravel the mysteries of the mind, so I started diversifying in the roles I picked. I got a postdoc and a little bit of funding to do some virtual reality (VR) exploration. Consumer VR devices were just coming onto the scene a decade ago, and we thought it would be a great tool for researching perception and so we did a project and we got a publication and that was all fantastic. We did the rubber hand illusion but in virtual reality and that was really cool. Some of this work also led to a short postdoc position developing VR-based educational experiments and resources of Macquarie Uni’s new cognitive science undergrad degree.
But my publication record was getting a bit diverse by this stage too and, in hindsight, I wasn't really thinking about that as much as I should have. I can see now how I really needed to be ticking all the boxes to stay employed in academia. Also, because I didn't start in psychology, working as a tutor didn’t feel like an option. So publications were really critical for me if I was to be a research-only academic
I did get all my PhD publications out, and an extra one, but one of these publications nearly killed me. It was the flagship chapter from my PhD; I felt really good about it, my supervisors felt really good about it, and we felt that I could make a really big, high-profile publication out of it. And if I got that big high-profile publication, then I'd be looking competitive for external funding, a DECRA or something like that.
(In case you’re wondering, a ‘DECRA’ is the Australian Research Council’s Discovery Early Career Researcher Award - a fellowship scheme established in 2012 that provides salary and project funds. Since establishment, success rates for this scheme have varied between around 12% to around 19%.)
This publication wasn't bad, but I was putting it forward in high-ranking journals and I kept getting back 'we can't accept this yet, do all these extra things'. And that's not an unusual story, I guess. But this was an 18,000-word publication, and so reworking that is a lot of effort. And I did that five times. I just went over and over this massive publication, and it was like doing another PhD. Eventually in late 2017 I got it published, but this happened at a time when there were a lot of other things going on in my life and redoing this massive publication over and over just wore me down.
I felt like I wasn't making progress, I felt like I couldn’t just keep doing this, I wanted to make enough money and contribute to my family, I wanted to move on and keep exploring and I couldn't. So, I just had to make the decision to move beyond academia. And I convinced myself that as a computational modeler, I don't need a lab. I just need my head and some time to read and to sit at a computer. And I can do that outside of academia. If I get a job in something like data science, then I'll learn even more skills and they'll be really useful as I continue research in my own time, and so I had this completely unrealistic idea that I'd go and get a job and then keep doing research on the side. And that's how I decided to move beyond academia.
JM: I did the exact same thing. I kept the honorary appointment at the university and had grand plans to keep publishing and I think there was one or two publications that came out after I switched.
I can definitely hear the push factors like the reasons why you want to look outside of academia for work. You briefly touched on this, but what led you to choose to go into data science work rather than another computationally heavy field?
SP: At the time, data science was a bit of a buzzword – I guess it is still, but not as much as it was back in 2016/17. And, like any good PhD student, I was just really good at self-education, I had started reading and had done a couple of online courses, and I realised that a lot of the skills I have fit this field. I think that applies to psychology and cognitive science, everyone has to learn a load of statistics and a lot of these skills are really highly transferable into something like data science. Particularly if you've got the coding background. At that time, a lot of statistics in psychology was being done in SPSS, so I had a bit of statistical knowledge, but I also had a coding background too and I thought this would be a field that I can transition into, maybe less on the data science side and more on the machine learning side.
I started applying for jobs and I found a startup that was looking to automate data analysis, to build AI to perform data analytics, and they were looking for a bunch of PhDs to really hype the profile of the company (I wasn't the only PhD that they employed, they employed a few). But I thought this was really cool. I'll be doing artificial intelligence and learning all these skills that I can then use in my cognitive science hobby research. In practice, I ended up being a software developer, because the realities of a start-up are we have to build what works and what sells. And on top of that, I didn’t have the time for much cognitive science stuff on the side given the demands of a full-time job, a long commute and wanting to conserve time and energy for my family. I’ve written hundreds of thousands of words, but it’s really easy and quick to jot down your ideas and much harder to craft that into a publication.
So I was doing this for a year and a half and then I think this company realised they needed software engineers, not self-taught coders that are really smart and looking to come up with AI approaches, and so they made a whole bunch of us redundant. So that was my first foray out of academia.
JM: That must have been a bit of a blow, especially having switched over and seeing that potential path opening up and then getting cut short.
SP: Yeah, I took it harder than I should have, but I think that's very normal. I'm in another startup now, and I've been there for over three years, so I understand that world a bit. And having gone through that experience, I can see how I took it too personally, and it wasn't anything wrong that I did. It's just that working in a startup is difficult. But I'd felt like I'd invested a lot in that job and so that was hard to be let go.
And I went for a few months where I had difficulty finding another job. I did some contract work. I was trying to be a data scientist, though by this stage data science jobs wanted something much more specific than my generic skills - what's maybe obvious from my background, but I'll spell it out here, is I'm a really good generalist: I'm able to go deep when I need to, which I did for my PhD, but I'm good at being able to cover multiple areas enough that I know what I'm talking about and can do a bit.
But these data scientist jobs were looking for a narrow depth of knowledge that I didn’t have, and so that’s why I wasn’t having much luck. And it was quite demoralising and my confidence really took a big hit. And when I eventually found a job, it was because I moved away from just trying to be a data scientist. Instead, I’d grouped up some of my other skills from this previous life in energy. I knew a lot about renewable energy, energy efficiency, and I have a degree in electrical engineering. And even though I'd spent a decade in cognitive science, I had kept roughly abreast of that world, so it wasn't foreign to me. And so I managed to get a job back in energy but doing data-heavy work.
It worked out really well because this company really needed someone like me. The company that I work for is a cloud software platform that orchestrates thousands of renewable energy systems, such as rooftop solar systems, into virtual power plants. A lot of the people that I work with are software engineers, but they need domain knowledge. They need someone that speaks software geek but that understands about renewable energy and I was almost the only one. A generalist like me was just perfect in this job. I could tell them everything that they needed to know about renewable energy. I could troubleshoot problems. I could still do a bit of coding and troubleshoot some of the coding problems. I could read code to look at what had been implemented algorithmically. The DevOps side is still pretty foreign to me, but I knew enough about how to use Git and GitHub and SSH and software engineering tools.
And in both of these startup jobs, I'd picked up enough product management knowledge to really contribute there as well. And startups just need this, they need people that can do their job and can also just jump on any problem that arises. This multiskilled background was really useful. Plus having the confidence to jump on to a problem thinking ‘I've done all these other things in the past, I know how to work through a problem just give me some time to do it’. So that’s made me really valuable in my current role. I’ve been happily there for three and a half years and it’s just such a good fit.
JM: Yeah, it sounds like you've found this perfect place where everything's kind of coalesced together.
SP: Yes, from a skills point of view, everything that I got out of cognitive science, on top of my previous life as an electrical engineer, just all lined up to make me a really valuable employee.
JM: A cynic might say, “alright, so he started in renewable energy, and now he's in renewable energy, what additional value did the PhD provide in this career path?” So I'm wondering: are there skills, experiences, knowledge from the PhD that you have brought back into this new role that's proved useful?
SP: Absolutely. If I had never gone into cognitive science, I’d be relying exclusively on advanced Excel. Learning about data science and machine learning, and being able to become a skilled coder, none of that would have happened if I hadn't gone to do my PhD.
And then there’s the problem solving that comes with doing a PhD, finding a way to make progress when it just seems like a massive task, knowing how to break it down into steps and get through it. The PhD for me is a source of confidence too: now I have the confidence to try some things that might not work out because I know that I did this amazing thing and it’s part of my bedrock self-esteem in a way.
I also learned that I can persevere – having persevered at the one PhD project for three, almost four years. Communication too – being able to write, summarise ideas, present my ideas in a way that’s influential for an audience.
All of that was honed during my PhD, I learned so much in my PhD, both skills- and attitude-wise, and it's all super valuable in my current job.
JM: It's been great hearing your career path and we really appreciate your candor in discussing some of the highs and lows in your career so far. Do you have any advice that you would give to someone who's considering moving from academia into another sector?
SP: I do, so much advice! I understand why I wanted to stay in academia, and why other PhDs would want that too. It might seem unwanted or just ridiculous to think about moving. But I think I might have sold myself a bit short, and I’ve seen others doing this too, in thinking that they have to accept this ascetic lifestyle and this hardship and this transience, needing to be ready to move anywhere in the world, just accept what you're given without question because you've decided to be an academic.
Instead, I think you need to have your ‘best alternative to negotiation’, because there is this alternative to academia out there. I don’t think it should be academia at all costs, or you risk looking back feeling you’ve wasted years of your life – if you're not getting the results that you want, then there is this other alternative to academia and it can be really good.
Also, if you do decide to move then your skills as a PhD can translate into earning potential. This is not to say that’s the end goal (let’s face it, if you’re working in academia you’re probably not motivated by making lots of money), but there's value in that money, even if not for yourself, but for others or family or even to donate to charity. And so I encourage postdocs to not ignore that value.
And I guess to back that up, I'd say that when you're in academia, it can seem daunting to consider moving beyond academia. But you just kind of have to trust the stories of people like me that the skills that you learn in your PhD absolutely will be transferable outside of a PhD. And you may want to take some little steps, like spending some time on job search websites and looking at the kind of jobs that are out there. Taking little steps like that starts to make it seem less daunting as you see there’s actually a lot of jobs out there and maybe you could do some of them and some of these skills do seem to match up with what you can do. And even if it's not perfect, you know that you’re smart, and you’ll learn what you need to learn and there absolutely will be lots of roles that you'll be able to do and be good at. It is a big change, but it's not so big that you can't do it.
Also, through doing your PhD you have this community, and that community doesn’t just disappear. For instance, I still speak with my supervisors from time to time. I recently reached out to my primary supervisor just to chat about where I'm up to and some of the things I've been thinking about but struggling to really turn it into something concrete. I bought him lunch, and we chatted and he gave me some great advice. But that community is really useful with missing academia once you’ve moved, if that’s something that happens to you.
JM: I think that's a fantastic bit of advice and a really lovely way to summarise, because you're right it can feel like there's the academic world and then this entirely discontinuous thing out there. And at least for me, the switch wasn't that traumatic.
SP: It's maybe adulting as well. I didn't feel like an adult until I was mid-PhD in my mid-30s, but there are just trade-offs you have to make as an adult. As passionate and as motivated as you are, sometimes you just can't do all of the things that you want.
I wanted to be a great husband and a great father and good to my parents in their final years as well as be an amazing postdoc and find the best jobs and it just wasn't working. It was a bitter pill to swallow in some ways, but it was my choice. And I haven't forgotten cognitive science, and maybe in a few years I'll decide that I've done this enough, and my kids are a bit older, and things are a bit more financially comfortable and maybe I can change things up a bit. There’s always changes that you can make.
And with that, we thank Stephen for sharing his story with us. We’re really feeling the power of change right now, and we wish him all the best for whatever might come next in his post-PhD journey!
We’re only a few interviews in so far, and while we’re busy writing these generously shared stories up for you, we will share that we’re already fascinated by the stories we’ve been hearing: so much has resonated, but so many remarkable differences between the stories too.
If you are a PhD who has moved beyond academia and you are interested in sharing your story, please get in touch with us via our LinkedIn pages - Sam, Jonathan - we’d love to hear from you.
Views, thoughts, and opinions expressed by persons interviewed in this series are solely that of the persons interviewed and do not reflect the views, opinions, or position of Sam, Jonathan or Sequitur Consulting.