Writing your dissertation in law (LLB and LLM)
For both the LLB and the LLM, the dissertation is a core component of the degree programme. In my experience advising students at both levels, I’ve picked up a few things along the way that I think might be useful to share. That’s the reason for this post and I hope it will be of some value whether you are preparing to design your project, currently researching or even writing up.
The topic
One of the great things about a dissertation is the intellectual freedom it gives you to explore a topic that genuinely interests you.
That said, your choice will usually need to fit within your department’s remit and the expertise of the academics available to supervise you. In other words, it would be hard for you to write a dissertation on the theory of quantum mechanics if you are in a law department. And as obvious as it seems, some departments may narrow the scope of your dissertation, particularly if you have chosen an LLM specialism, or depending on the staff available to provide you with an adequate level of supervision. My advice? Have more than one topic in mind. Being flexible can save you a lot of time and frustration down the line.
It is commonplace, particularly in departments with strong international communities, for students to pick a topic related to their home jurisdiction or country of origin, and that is perfectly doable. Just bear in mind that you will need to make the topic easy to digest for a reader who is not versed in that area of the world. It is also the case that so-called sexy topics become very trendy. As a legal scholar who supervises dissertations, I often see AI in the titles. Years ago, it might have been the metaverse, bitcoin, non-fungible tokens (NFTs) or the Internet of Things. In international law, students are also naturally drawn to current conflicts and other major geopolitical developments.
Is there anything wrong with choosing a trending topic? Not at all. Current issues can make for excellent research. But a timely topic isn’t, by itself, a strong research project. You still need to explain what makes your particular angle interesting and what argument you’re making. In other words, don’t just tell your reader what you’re writing about; show them why your approach matters.
The core argument and why it matters so much
A dissertation can tackle a timely topic. It can have very good sources. It can even be very well written and show a strong knowledge of the law relevant to topic X or issue Y. But if the dissertation is not advancing a claim, something is not quite right. From my experience supervising students, reading drafts, and marking final submissions (of my own supervisees and other students too), I’ve found that dissertations with a clear central argument tend to be much stronger. They have direction. The reader can see where the research is heading.
Many supervisors start with the topic and move straight to the overarching research question, chapter structure, methodology, etc. That’s a perfectly reasonable approach, but I tend to do things a little differently.
I start with the topic and the problem. I ask students to explain, in their own words, why the issue deserves legal interrogation in the first place. What is the problem? Is there a tension, a gap, an inconsistency, or something about the current legal position that deserves closer examination? This is not easy question and takes a couple of weeks to kick in but it resolves as the student reads more about the topic. Then, before moving on to the research question, I ask a different question: What argument do you think you could make about this issue?
Why? Because without thinking about the argument early on, it’s easy to end up with a research question that is perfectly understandable but doesn’t really lend itself to a critical analysis. For example: What laws apply to AI-generated music? or To what extent does the law apply to algorithmic profiling on social media?
Those questions aren’t necessarily bad starting points. But on their own, they can encourage students to describe the law rather than critically examine it. Identifying the relevant legal framework is important, of course, but a dissertation needs to do more than tell us what the law says. It needs to make a case for something. My approach encourages critical thinking from the outset and gives the research a clearer sense of direction.
So, if you’re planning your dissertation right now, try asking yourself two questions: What is my argument, and what is the story I want to tell?
You don’t need to have all the answers at this stage. You might have several possible arguments in mind, and that’s absolutely fine. Think of them as provisional claims that you’ll test as your research progresses. You’ll need to examine them against case law, legislation, policy developments, and the relevant academic literature. Some claims won’t hold up and it’s fine to let them go. Others might need to be refined or combined. That’s all part of the process.
The argument you settle on shouldn’t necessarily be the one that best reflects the prevailing academic opinion or the one you find most appealing at the outset. It should be the one you can defend most convincingly in light of the available evidence and the strongest objections to your position. The goal isn’t to find an argument that no one can challenge. It’s to develop one that can withstand serious scrutiny. One thing that we supervisors and markers appreciate in a dissertation is its intellectual honest about its limitations and counterarguments. A weak dissertation pretends that the weaknesses in its argument don’t exist; while a strong one shows that the author has thought carefully about them. Once you have a clearer idea of the argument you want to make, developing the research question becomes much easier. Rather than trying to invent a question in the abstract, you can work backwards from the claim you want to investigate and defend.
The research question
It is not unusual for a dissertation to try to tackle several questions, because the topic is timely and the student wants to research several areas around the same problem. If that is your case, it is doable, but it has its limits, and you may end up down a rabbit hole if you don’t govern your curiosity. So you need one, and only one, overarching research question. I like to think of it as the circulatory system of your dissertation. It should run throughout the entire project, connecting the different chapters and keeping the whole thing alive. Is it the entire body? Of course not. But without a functioning circulatory system, the body wouldn’t survive. Your dissertation works in much the same way: the research question gives it direction and coherence, even when you’re exploring different aspects of a broader problem.
Once you have that overarching question, you can break it down into smaller sub-questions. These will help you organise your research and structure your chapters.
What makes a good research question and sub-questions?
There’s no foolproof formula for a good research question, but in my experience, questions framed around ‘how’ often give students more room to develop an analytical argument than questions framed around ‘what’.
Let’s return to the examples from the previous section. Imagine your research question is: What laws apply to AI-generated music? or To what extent does the law apply to algorithmic profiling on social media?
These questions can be useful starting points. The problem is that you might end up producing a comprehensive account of the law without developing a particularly strong argument of your own.
The reasoning is relatively straightforward. A what question often asks you to identify something, and the answer can end up being a list. A how question, by contrast, invites you to explain how something works and to reimagine parts of the legal system. Those explanations can be pushed and tested. And that’s where your argument begins to take shape.
But, as I remind my students, “how” is not a magic word. How does the law regulate AI-generated music? would not do considerably better. What makes a how-question analytical is that it pairs the law with a tension or a consequence. By contrast, How should section 9(3) of the Copyright, Designs and Patents Act 1988 be reformed to address AI-generated music? would allow you to explore the issue in much greater depth. Another example is How can social media platforms adjust their data-retention practices to respect users’ right to erasure under the GDPR? Again, the question gives you a legal problem to investigate and an opportunity to develop a reasoned position on how it should be addressed. What matters is the legal problem you’re asking the question to address.
Now that you have a research question, you can start thinking about sub-questions. I always suggest no more than three. These should derive from the main, mother question, and between them they should cover everything you need to answer it. Here you can use what-type formulations, along with when-type and, of course, more hows.
Example 1: AI-generated music and copyright
Overarching research question:
How should section 9(3) of the Copyright, Designs and Patents Act 1988 be reformed to address AI-generated music?
Possible sub-questions:
- What does section 9(3) currently provide for, and whose interests does it protect?
- When, if at all, should AI-assisted music be treated as a computer-generated work rather than a work authored by a human?
- How might approaches adopted in another jurisdiction inform potential reform of the UK legal framework?
Example 2: Social media, data retention, and the right to erasure
Overarching research question:
How can social media platforms adapt their data-retention practices to ensure effective respect for users’ right to erasure under the GDPR?
Possible sub-questions:
- What does Article 17 GDPR require of social media platforms, and what exceptions apply?
- How do current data-retention practices create difficulties for the effective exercise of the right to erasure?
- How could the principle of data protection by design help platforms develop data-retention practices that make the right to erasure more effective?
Of course, these are starting points, not templates you need to follow mechanically. Depending on your topic, you might need a different structure. What matters is that your sub-questions work together to answer the overarching research question and, ultimately, support the argument you want to make.
Getting this right early on can save you a considerable amount of work later. It also gives you a much clearer idea of what belongs in your dissertation and… just as importantly, what doesn’t.
Methodology
For undergraduate law dissertations, a doctrinal approach is probably the most common starting point. And that makes sense. Most LLB students have had limited exposure to empirical or interdisciplinary research methods during their degree, so doctrinal methodology is usually the way to go.
But a doctrinal dissertation does not have to be a dissertation without literature.
One of the most common areas for improvement I come across is work that makes very good use of cases and legislation but pushes the literature, the academic voices that give meaning to those legal definitions and cases, to the margins.
As a non-doctrinal scholar myself, I have great respect for colleagues who teach and publish from a doctrinal perspective, and I have learnt a great deal from them. Above all, I have learnt that the most sophisticated doctrinal scholarship goes one level below the text of the law and the judicial decision to uncover the rationale behind them. The most celebrated judgments, the ones that populate the syllabi of tort or contract law around the world, are often those that engage with ideas from philosophy or economics, precisely because positive law absorbs and crystallises those ideas.
Likewise, doctrinal articles in leading law journals are unafraid of drawing on non-legal insights to offer a richer interpretation of the law. Your dissertation should use the cases and statutes to establish what the law is, and the literature to explain why it is that way or whether it should be. In the LLB, you are not expected to engage directly with sources from other disciplines. But if you take a closer look at the legal commentary on your topic, you will very likely find that scholars have already placed their doctrinal analysis in conversation with philosophy, political theory, economics, history, etc. That’s your way in. Cite them, bring their ideas into dialogue with your own argument, and you will expand the reach of your analytical voice without leaving the discipline of law.
In the LLM, there is a bit more room for non-doctrinal analysis, as some students arrive after a few years in practice or with a previous degree in another discipline. If that sounds like you, you may have more tools at your disposal, and you’re welcome to use them. But you need to explain them clearly and justify why they suit your question. In my personal experience as a researcher in law and technology, I find that a strictly doctrinal reading of the law is rarely enough to meet the challenges that technology poses, and other angles are often needed, such as concepts drawn from science and technology studies (STS) or political economy. Would that work for you? Not necessarily. Choose the methodology you find most useful, but you need to announce it in the introduction and justify your choice. Talk to your supervisor early. They may not always be familiar with the non-legal framework you have in mind, but they are likely to support you if you make a convincing case for it. In fact, we supervisors are glad to be introduced to new areas of knowledge, and even new methodologies, by our students. It is refreshing, provided it is done with rigour and care, and with a clear commitment to using the framework throughout the dissertation. What does not work is announcing an interdisciplinary approach in the introduction and then leaving it behind as the dissertation settles into an IRAC structure: issue, rule, application, conclusion.
If a doctrinal analysis is what you really want to do, that is perfectly fine, but don’t dress it up as interdisciplinary; the mismatch will distract both you and your supervisor. You are better off adopting the approach suggested above for the LLB. It is not unusual for an LLM dissertation to achieve a distinction with that kind of serious, reflective doctrinal methodology.
Structure
Your dissertation must, without exception, have an introduction and a conclusion, with a number of substantive chapters in between. I usually suggest four, but some of my students have written five or six and it has worked well for them. Again, there is no magic formula, but this is where the research question comes in again.
Remember the circulatory system analogy from the previous section? Your overarching research question runs throughout the dissertation, but it’s your sub-questions that help you decide how to organise the chapters. As a starting point, if you have three sub-questions, I would suggest three substantive chapters, one for each, followed by a synthesis chapter – all of that along with the introduction and conclusions makes 6 in total. But don’t treat this as a rigid rule. You might need to spread a sub-question across two chapters if it involves too much material for one, or if a particular aspect of your research (a historical dimension, for example) deserves a chapter of its own. The important thing is that the structure serves your argument, rather than forcing your argument to fit an arbitrary structure.
The synthesis chapter is not the conclusion. It is the chapter that brings the three previous ones together and turns them into the findings, recommendations/proposals of your dissertation. This is where your argument is fully assembled and where your own voice should be at its strongest. And again, this is not the conclusion, because the conclusion is meant to do something different. It steps back from the detail and answers the main research question directly and concisely. What the conclusion should never do is introduce new material, new cases or new argument. Don’t try to impress us with complicated language. Impress us with the quality of your thinking and the clarity of your train of thought.
Communication and language
Lastly, if there is one sacrifice you should make, it is to give up complexity for clarity. A dissertation can be sophisticated without being hard to read. I am not talking about legal terminology here; that is fine and often necessary. I mean words and sentence structures that do nothing to help convey your ideas. Part of the blame, I suspect, lies with the spread of AI writing tools, which students use in the hope of sounding more professional and precise. Too often, the result is flamboyant prose that looks impressive from a distance, but once the reader tries to piece the ideas together, it conveys very little. Sometimes, it says practically nothing at all. Each institution has its own policy on AI, so familiarise yourself with yours and follow what it requires. But even if your institution permits AI use broadly, don’t let a chatbot dilute your argument. Your voice is the one being assessed. Honestly, one of the most satisfying things about marking a dissertation is reading a piece of work that is clear and well organised its argument. Don’t try to impress us with complicated language. Impress us with the quality of your thinking and the clarity of your train of thought.
Equally so, organisation is a matter of guiding your reader through your argument. This is what signposting does. Remember that your introduction is your GPS. Signposting is that same GPS speaking to the reader along the way, so they never lose track of where they are. Each chapter should open by telling the reader which sub-question it addresses and how it moves the argument forward. At the end, briefly summarise what you’ve established and explain how it connects to the next stage of your analysis.
The same principle applies within chapters. Don’t assume that the connection between two ideas is obvious simply because it’s obvious to you. Use clear transitions between sections and paragraphs so that your reader can see how one point follows from another and link it back to the main research question. Every time you link a chapter back to your research question and your main claim, you remind the marker that each part of the dissertation is doing work for the whole.
Throughout this guide, I’ve emphasised the importance of having a clear argument, but having a good argument is only half the job. Having a strong argument matters, but so does making it easy to follow.
Dr Mauricio Figueroa,
Assistant Professor of Law, Durham University – October 2026