Pathfinder Career Narratives 65: Linguist

Pathfinder Career Narratives is an ongoing series tracking the career choices and experiences of doctoral graduates. You can see all of the posts in the series here. You can find all the Pathfinder resources and opportunities here. Today’s blog is written by Dr Nathaniel Miller, who is a linguist for Reality Labs at Meta. You can find Dr Miller on LinkedIn

Name: Nathaniel Miller

Doctorate subject area and year of completion: Arabic Language & Literature, 2016

Role and employer: Linguist, Reality Labs at Meta

Shortly after being asked to write a career narrative, I received an offer for a contract role as a linguist at Meta, so this has become a narrative about how I got here over the course of about two years. To anticipate the moral of the story, I had to figure out what I wanted to do, instead of just what I didn’t want to do (more lectureships and post-docs).

For most of my life, my non-academic work experience was limited to haphazard semi-jobs like working after college as the office administrator for a family-owned shed company or teaching at NGOs in Egypt, where I lived for three years. I have a BA and MA in English and a cocktail-party-conversation-killing PhD in medieval Arabic literature. I was one of those people who thought solid scholarship could save someone from the job market. I taught, but I was much more research-focused. I held out for seven years in academia via post-docs at Cambridge and NYU Abu Dhabi. I also wrote a book.

Aside from my academic and non-academic work experience, I took an online course on data analysis during the pandemic lockdowns, so I do have a basic-to-intermediate grasp of Python, SQL, and data visualization tools. I also did some workshops on Natural Language Processing (NLP), so I have a high-level understanding of that. But other than that knowledge, some old song lyrics, and my understanding of shed construction, 95% of the space in my brain is taken up by Arabic.

None of the above was any help on the job market when I decided in early 2023 that I couldn’t stand the endless post-docs anymore. I recommend spending a maximum of three years on those, by the way. I didn’t switch at a great time. Tech hiring, for example, was two to three times higher in 2020 and 2021 than today. By the time my second post-doc was expiring in the summer of 2023, the pandemic-era hiring boom had subsided drastically.

When I started my search, I was thinking in terms of “translating” my soft skills into career paths: writing skills (UX, content writing, technical writing), research skills (business analytics, data analysis), and sometimes presentation experience from conferences and teaching (behold, a PowerPoint of business slides!). I applied for jobs in all those areas. I was very much thinking I would accept anything that paid well (I was targeting $80k-$100k), and I didn’t put much stock in what I actually wanted to do, just what I thought I would be able to get.

Aside from informational interviews (about thirty or forty), I worked with a career coach ($300 for two sessions and a resume re-write) and read things like Chris Caterine’s Leaving Academia, which I highly recommend. I spent about eight months searching for a job, five while still doing the post-doc in Abu Dhabi, and three while unemployed in Colorado. Compared to academia, the good thing was that there were a lot of jobs to apply to, and I got interviews. I applied for about three hundred jobs, and I got some sort of response (interviews, requests to complete tests or projects, requests for further materials, etc.) to about fifteen. It only took me a month of applying to start getting interviews. I interviewed, e.g., to do anti-money laundering content writing, to do customer-facing business software documentation, and to create social studies content for an ed tech.

In general, other than mastering the basic formatting for non-academic resumes, I did not find that tinkering with the resume or getting referrals made much of a difference to the interviews I got, aside from the obvious importance of tailoring a resume to every job, but I recommend spending as little time as possible on that (use ChatGPT). At my peak, I could apply to at least one job an hour. However, I had a hard time advancing past a first or second interview. Part of my problem was just learning to interview, but part of it, I think, was that my resume looked too academic, and I confirmed that in interviews by lacking clarity and conviction in what I was trying to offer. Because I would take anything, I felt I was a bit adrift and either couldn’t find the right catalyst to combine my disparate skill set or was still missing some crucial skill. I sensed that I wasn’t targeting a small enough niche, and I also felt I needed to do something else to stand out.

I finally found a job a step above entry-level doing labor law compliance investigations for the Colorado Department of Labor. It was fully remote, and I loved my colleagues and the work culture, but I was not happy with the compensation (Denver is expensive now) or the room for growth (the best policy opportunities went to lawyers). One reason I got this job is that government hiring is (or was) usually a lot less biased towards things like age, career stage, and background skills. It was acceptable to be “overqualified.” Another reason, I found out later, was that an ex-academic on the interview committee had spoken up for me—this was quite a common recurring experience I heard about from ex-academics I spoke to in informational interviews.

Having a full-time job slowed but didn’t put a stop to my search. After about a year, I started moonlighting part-time doing data annotation for an AI startup. I found the job through a simple job search; the description happened to mention “Arabic,” a term I would throw in the search rotation every now and again. A lot of data annotation can be poorly paid, transactional, and anonymous, but this one was moderately well-paying and relatively small-scale, since it was a startup. Data annotation is also quite boring, but I slowly developed more of a sense for how LLMs work. More importantly, I had the credential on my resume. The job helped financially, but it was not a long-term contract, and the amount of work wasn’t sustainable for me.

A couple of months ago I finally achieved some partial clarity. I had not really been thinking about what I wanted to do, in part because I found it hard to visualize where I was going. I needed to return to myself. I had been trying to get away from language as an unsought-after humanities skill, but I am basically a language guy. Despite my best efforts, every job I’ve gotten an offer for (four at this point) has had “Arabic” in the job description. The best professional relationships I have are with former language-focused PhDs and other foreign language people. Plus, I just find language interesting, and I had had enough boredom doing labor law and data annotation. I hate being bored.

Having to work directly with large language models for multiple hours every week sparked an occasional personal interest of mine, linguistics. I brushed up on some Arabic dialect grammar and bought some books I’d been wanting to read. Demand for linguistic expertise is high in the AI space right now. While it’s true that my PhD isn’t in linguistics, I do have significant experience with it, and a little academic experience goes a long way in the private sector when you have what they’re looking for—most academic research is major overkill compared to industry applications. More importantly, a key reason to apply to any job is to grow and learn more, so I knew if I got something doing linguistics it would be challenging and interesting.

One day I decided to just perform the job I wanted to get, something doing linguistics in the tech space. I changed the vestigial “Arabist” I’d always kept in my LinkedIn headline to “Arabic linguist” (I also dropped the “labor law compliance” and left in “AI” because the two together was jarring, I suspected, for recruiters). It felt like the perfect combination of authenticity and in demand. I think now that’s what you want to aim for — neither completely reinventing yourself nor trying to sell your obscure academic niche in the private sector. It’s hard to think about what you want when you’re unemployed or in an unfulfilling role. Anything seems better. But you will actually come off to employers as far more compelling if you put the horse before the cart and can show them you know what you’re looking for, what you’ll enjoy, what you’ll be good at, and what value you can add for them.

After I had real AI work on my resume from my data annotation contract role, I started hearing from recruiters on LinkedIn and Dice. One reached out about a contract role for a linguist at Meta, working on automatic speech recognition for their smart glasses in the Middle East market. I was interviewed by a PhD linguist for this job (pro tip: if, like me, you get nervous in interviews and don’t naturally emote enthusiasm, at least say you’re enthusiastic about the role and why) and a linguist friend who also works in tech provided guidance and perspective during the interview process; draw on that network of ex-academics! No magic bullet landed me the role (I didn’t get a referral or even tailor my resume), but introspection, networking, and my side hustle all gave me something substantive to say in the interviews. It just felt like an incredibly good fit and that’s what I’m doing now.

My post-academic journey has been as important as the destination and to a large extent the journey shaped the destination. It took a couple of years, I went through one full-time job that didn’t seem like a perfect fit, had a part-time contract role to bolster my resume, kept talking to a lot of people, and had good luck (Meta wanted someone who could do a hybrid work arrangement and I’m near a Facebook office). I didn’t figure out what I wanted to do in the abstract; I learned (by doing) more about what was out there and connected that to my own interests and skills. Since my current role is also contract, I know that the journey isn’t over yet, but I’m generally happy with the direction it’s going in.