Skip to main content

How to write a thesis

Your thesis is like your first love: it will be difficult to forget. In the end, it will represent your first serious and rigorous academic work, and this is no small thing. - U. Eco

A Thesis in Theoretical Physics

You visited an advisor and got a topic to work on in theoretical physics, congratulations! Now the only thing you are left to do is study, do the research, wrap it up, and write it down. As for the Tools of the trade article, this list has a down-to-earth approach on providing a pragmatic look at tools and advice regarding your thesis. As in other articles, I will be as general as possible and as specific as needed. I will describe my suggestions for doing a thesis in general -> in physics -> theoretical physics -> theoretical nuclear physics -> and my Lund group in particular, both at undergraduate, graduate, and PhD level.

There are entire libraries, websites, and initiatives dedicated to the craft of writing in general and academic writing in particular. Nice initiatives and tools on general writing are shut up and write and Hemingway App. There is also plenty of material to take inspiration regarding academic writing. Most interestingly, there is a whole 300-pager by Umberto Eco: “How to write a thesis” (here you can read the review and excerpt). Online you can find the book if you wish, but don’t waste precious thesis time (this post is already more than long enough). Keep in mind that Eco’s book was written in the context of Italian humanities, where a thesis lasts easily more than a year of pure writing; therefore, it is more applicable to a PhD thesis than an undergraduate thesis. Lund University (LU from now on) has its own resources on academic writing. There are courses, workshops, and an interesting website.

Learning to have strong academic writing is a lifelong endeavor. It is not possible to master every process at any given stage of your studies. However, by following advice and practicing, you will become better and more confident in your writing.

Context

The thesis is the final academic document testifying to some work required for the attainment of a degree. There are theses for bachelor's, master's students, licentiate, and PhD degrees. Theses are used even for some professional or professor habilitation in some countries and circumstances. Therefore, even though topics, length, and depth might differ from thesis to thesis, they always have the same primary audiences: the people handing out the degree. In LU, the B.Sc. and M.Sc. graduation theses are refereed by one or two external examiners. In our case, they are usually people in mathematical physics who have experience in many-body systems but not necessarily in your method of choice or nuclear physics.

When writing anything, the first thing to keep in mind is the reader. Like your examiners, other people who might read your thesis are knowledgeable about the field but not of the specific argument. For example, in our case, they will be your fellow students who might need to pick up your work. That is, prospective physicists but not necessarily with a nuclear or theoretical physics background. You can take for granted that the reader knows what a Lorentz transformation or quantum state is, but you should not abuse field-specific jargon and use it without introduction. Every acronym, method, and code must be introduced and referred to with references.

The use of references has to be strategic. Being the thesis an official document for the attainment of a degree, it has to “stand on its own feet.” The reader from your target audience has to be able to read comfortably without the need to constantly refer to the literature. Of course, you need to use references and cite relevant literature, especially to provide plenty of examples and material to study in more depth. However, within reason, everything you use for your results needs to be introduced explicitly so that the content and context of your work is clear.

The work done for a thesis in physics is usually centered on research, either by critically reviewing previous research results or by developing original research guided by the supervisor. Bachelor's and Master's theses are 15, 30, or 60 credits, corresponding to 10, 20, or 40 weeks of full-time work, respectively. The goals are usually set by the supervisor, and the amount of supervision and independence will depend on the specific project and be adjusted according to performance.

Last and probably least, another consequence of being an official document is that the thesis often has to adhere to some official or unofficial guidelines, usually concerning length, structure, format, and, rarely, content. For the Lund physics department, you can find the guidelines here and here. Here is the checklist for the registration of a Physics diploma work in LU. Pay particular attention to the learning outcomes.

Reading

Bibliographic Research

The first thing when approaching thesis work is to understand the scientific background and context of your work. This is done by reading articles and books suggested by the supervisor that are instrumental to the problem. Some articles are worth reading and understanding in detail; others, to skim to grasp the main concepts and results. Only experience can judge how much to devote to each article and how to read and understand effectively. It is not an exact science but an art that improves with experience.

Your thesis work is the opportunity to delve into the literature and start to gain this experience, picking the brain and experience of an expert supervisor, so make the most of it. Try to read academic literature every day. Read everything that you think is worth citing and everything you will cite in your work. Read modern developments in journals and the arXiv of your field. It is not uncommon for a thesis work to review dozens and even a few hundreds of articles. The articles your supervisor cites are only the starting point of a journey of understanding.

In the writing of your thesis, especially in the introduction, you will need to refer to the literature in order to point the reader in providing context and pointers to concepts and tools you used in your work. In the same way, scientists use references in articles and often in books. Therefore, you can use the bibliography of the article you read as an important tool for your bibliographic research. You can follow citations in two ways:

  • upstream, looking at an article's references to understand on which other works it is based;
  • downstream, looking at works that cited the said articles and use it for follow-up works.

This is crucial to understand the scientific foundation and impact of a work.

At LU, a short training course is given in Language and Library.

Textbooks, Journals, and Articles

There are different outlets for scientific publications. Textbooks are published by a publisher. Articles of different types get published by a journal. Topical journals are the traditional and always-good way to read and update about new results in a field. The editorial collocation of an article is an indication about its subject, novelty, and mesdian impact. Unofficially and roughly, they can be categorized in the following way:

  • Textbooks: You encountered textbooks in your basic education. Academic textbooks are often more advanced, but they are written to be a comprehensive, reliable, organized, and pedagogically useful treatment of an argument. There are few updated books in nuclear physics; in later years, the community has been relying more and more on articles.
  • Review papers: These are long overviews of an argument published in a journal. More updated, limited, and cutting-edge than a book, they may contain new results. They are a good starting point to work on an argument, especially if books are not available. Journals publishing reviews are, e.g., Review on Modern Physics and Reports on Progress in Physics.
  • Articles: These are the “standard” scientific publications, describing new results in as much detail as needed for understanding and reproduction. It is good practice to periodically read issues of the journals publishing articles in the field you want to stay updated in. For physics, good resources are the APS journals, in particular, Physical Review C (shortened PRC) for nuclear physics and Physical Review E for many-body systems and non-linear phenomena. In these journals, some particularly interesting articles get featured on the homepage as editorial suggestions.
  • Letters: These are short articles used to communicate particularly novel and timely results that the community should quickly take notice of. For this reason, on average, letters have a higher impact, and the selection is often stricter. Topical journals like PRC have “rapid communication” sections for letters. Letters are often targeted to a wider audience of physicists and even scientists in general. Being featured in Physical Review Letters (shortened PRL), Nature, and Science is an achievement for any physicist.

Bibliographic Tools

To organize the work of bibliographic research and citation, apart from the quite important brain and the internet, it is sometimes useful to be helped by tools:

Some people use Mendeley, but I don’t feel right endorsing bibliographic options owned by editorial companies.

Working

This will probably be your first experience in original scientific work. Arguably, your objectives should be:

  1. To learn as much as possible.
  2. To do a good research job that feeds into the primary objective.
  3. To present it properly. That is part of the learning outcomes for the diploma work.
  4. To think about the role of science and your work in business, society, and your future.

Here is the list of learning outcomes for the diploma work of B.Sc. and M.Sc.. These are no small technicalities but set the expectation of the quality of your work required by not only LU but the ministry of research and education. Be mindful of the responsibility that the title you are applying for carries.

To organize the work according to these requirements, you have to coordinate with your supervisor. Set a timeline and schedule. Keep in mind that even the most open and available of the supervisors is probably a busy person and has other duties to attend to and frequent trips. Be sure that he is available for any strict bureaucratic or work requests you have from your project.

Time management is your responsibility, and being open about the duties and requests you have is an important part of efficient project management and hence successful work. Check the deadlines and appointments. According to the type of work and credits you have for the project (1 credit is 25-30 hours of work), the workload will be set accordingly, and the supervisor will help you set realistic goals.

Programming

Some research requires coding to simulate and understand the physical system and formalism. The tools of the trade article can help you find some tools and resources. Regarding the context of the thesis work, one word of advice is: do not try to do it all. Choose a few tools to perfect and focus on getting the most done and being effective for your project.

To help the organization of the work and collaboration, it is sometimes efficient to use Git. For this reason, at the division of mathematical physics, we set up our own GitLab server (not to be confused with the public gitlab.com). Focus on the objectives and structure the code accordingly.

It is good practice to use Git as a versioning system (not anymore v1, v2), and when you get the hang of it, it is convenient to use it also for important documents, such as the thesis.

Writing

Scope, Tone, Language

The tone and language of the thesis have to be gauged according to the objective and the audience. The audience are your examiners and your fellow students. You have to write for prospective students who need to understand the scientific context, have a good bibliography to start from, and a report of your results useful to reproduce and continue your work. Even more than usual, write only what you really know to be correct. Typos happen. Imprecise concepts, incorrect statements, wrong equations will not help your reader, and therefore you.

Scientific prose has to be crisp and precise. Use short and clear phrases. Keep the grammar simple and exact. Choose your words precisely. The objective is, first and foremost, a dry, correct, and objective account of your research and results.

A modified version of George Orwell’s rules for writing can be used: > A scrupulous writer, in every sentence that he writes, will ask himself […]: What am I trying to say? What words will express it? What image or idiom will make it clearer? […] I think the following rules will cover most cases:

  1. Never use a metaphor, simile, or other figure of speech.
  2. Never use a long word where a short one will do.
  3. Without compromising precision, if it is possible to cut a word out, always cut it out.
  4. Never use the passive where you can use the active. Use the first person singular, when is work you (and only you) have done. Use the first person plural to refer to the group or the community. Use “One” to refer to an eventual reader. Use the passive voice when needed, especially to refer to the work itself
  5. Never use a foreign phrase, or a jargon word if you can think of an everyday English equivalent. Use the scientific words respecting their context and meaning
  6. Break any of these rules sooner than say anything outright wrong.

In addition,

  1. Equations are part of a phrase; use punctuation when introducing (not : but ,) and after the equation (usually , or .).
  2. I cannot stress this enough: define everything you use. Every symbol and index in an equation, quantum number, content of a figure, axes of a plot… etc… Attach captions to figures and tables.
  3. Refer to equations as Eq. (*). Figures as Fig. *. Tables as Table *.
  4. Write, both thesis and code, for yourself in the future, when you will have forgotten what that index in the third line of equation (7.24) was about.

If you read as suggested, you will pick up the style of your discipline. Try to imitate it.

For more information, a short training course is given in LU regarding Language and Library.

Constraints

Being the thesis an official document, it is extra important to respect official rules. One of the most relevant regards plagiarism. Literal quotes of other works have to be in quotes and properly referred to. Non-original figures also have to be cited, even when the copyright is available and free to use. Plagiarism is a serious offence and can ruin careers and lives. LU has a zero-tolerance policy on plagiarism on diploma works, including self-plagiarism (copying one’s own work). To guarantee this, all theses are passed through a plagiarism detection system called URKUND. Submit the thesis to URKUND a few days in advance of the deadline.

The number of pages of a report varies enormously according to topic and originality. A research thesis requires fewer pages than a review one. At the Physics department of Lund, a (somewhat) strict limit of pages for diploma works is in place:

  1. 15 credits B.Sc. report: 25 pages max;
  2. 30 credits M.Sc. report: 40 pages max;
  3. 60 credits M.Sc. report: 50 pages max.

This can also work as an indicative size for similar works.

Other constraints might be in place, depending on your field, University, and situation. Formalities such as a cover page are often required. Moreover, Lund’s Physics Department also imposes the sections that have to be present in a thesis.

Structure

The title of the thesis should illustrate the work you have done. There is no point in too general titles (“Nuclear physics”); too specific titles (“Study of 2+ states in rotational bands using HFBTHO code in the Praseodymium isotopic chain”) on the other hand discourage the reader that might be interested in more general concepts. As with many things related to writing, you will have to strike a balance. Let’s use the latter example to guide you through the process, considering you evaluate this to be your contribution. Your study might not only be interesting for people looking for 2+ states. For sure, if your study is in physics, the results should not depend on the code used. Hence, without loss of information, “Study of rotational bands in the Praseodymium isotopic chain” is definitely more useful for people who need to decide if your thesis deserves a second look.

When writing, you should always ask yourself what is needed here, why, and how is it possible to improve it, especially for important sections like the title and abstract.

The abstract is a short summary of a few lines. It regards the premise, main method, and results and conclusion of your work. A thesis summary is not much different from an article; therefore, you have plenty of examples at your hand.

In the appendix of the diploma work are specified the necessary sections and content of a thesis.

If you allow me a kitchen metaphor, consider the thesis as a hamburger: the Introduction is the restaurant, table, and plate; the Method the bottom bread; the Results the patty; the Conclusion and Outlook the condiments; the Bibliography the top bun; the Appendix, code, and other documentation your complementary fries and beverage.

Introduction

The introduction is the support and presentation for your work. It is needed to introduce your work and its scientific context. Use what you have read, but don’t exaggerate with background information. A thesis is not a textbook. The main objective of having a context is to introduce the significance of your work, why you are doing what you are doing, and how this helps the scientific community. One of your student colleagues should be able to be introduced to the topic, have the pointers to the literature needed to understand deeper, and be compelled to continue reading.

Method

The method section is the foundation of your work. It is not strictly required by the syllabus and can eventually be merged with “results.” However, it is good practice to keep them separate. Here, you should introduce the techniques that will be used in the results section to decrease the reliance on external reference material and make your thesis self-sufficient.

For example, the Hartree-Fock method or cellular automata are examples of well-known techniques that might be needed to understand your work. A brief and to-the-point description of this well-known method will help the reader. But restrain yourself and describe only the methods that are most relevant to your work. Other background information should be referenced to the literature. Remember the page limit and to preserve the sanity and disposition of advisors and examiners. Think that we have to read a few of these theses in a week, and while we want to verify that you understand the concepts, reading pages of well-known, irrelevant details does not put us in the mood for a positive evaluation.

Results

The results section is where “the beef” is: the main content of your work, your original contribution. Here, you use the methods introduced within the scientific context explained in the introduction to provide new insight into the topic of your thesis. Depending on the type of thesis, stage of studies, ambition, field, it can be radically different. The results section is the one most comparable to articles. Therefore, you should take inspiration from the literature on how to present your results.

Here more than ever, you have to consider Orwell’s suggestion: ask yourself, “What am I trying to say? What words will express it? What image or idiom will make it clearer?” Try to focus a message and think of the best way to convey it.

A common mistake is thinking of the thesis as a simple laboratory report, where you are tempted to list all your trials in chronological order. Introducing results chronologically might be an efficient strategy (often a thesis progresses in complexity and builds on previous results), but it is not always the best strategy. Focus on the scientific message, and select those results that are important to illustrate that message.

Conclusion and Outlook

The conclusion gives the flavor and aftertaste. What do you want the reader to take away and remember? What are the discoveries you made in your work, and how do they fit with and contribute to our understanding?

Moreover, an outlook must also be provided. That is, suggesting possible avenues for continuing the journey you started. What should we do next? Why?

Bibliography

The well-researched and redacted bibliography is an essential part of a text. It provides both motivation, context, and possibility to investigate deeper. In good bibliographies, you can find insightful texts and hidden gems. An expert examiner (or referee) can almost judge the quality of a work by only looking at the attached bibliography. The bibliography is a good marker of quality because it is a marker of the intellectual “diet” of a person. The more varied, deep, and sophisticated the diet, the higher quality the work will usually come to be. An intellectual is just as good as his/her reading list, and scientists make no exception.

Curate your reading list and demonstrate good use of the bibliography. Readers will be grateful.

Appendix and Others

The appendix is an additional part of the text. It is a good and sometimes necessary addition. Interesting derivations, ancillary results, and additional content can enrich the text and provide details for the not-so-average reader. In the main text, you target the audience of examiner and fellow students who need to understand the scientific contribution you made. The appendix will be reserved for the reader that wants more details, the student that has to pick up the work, someone that might want to implement something you derived, who wants to know the nitty gritty of your results to reproduce them.

Before my time, way back when dinosaurs roamed the earth, codes used to be attached in the appendix. Today, it is not that useful to have a line-by-line printout of the code. It is way easier to provide a link to a public or semi-public repository (like the division’s GitLab), and often codes are now too complicated to be printed out with ease. However, this is an excellent example of the content of an appendix: something perhaps not directly scientifically relevant but informative for people who want to look closer and work it out for themselves.

Tools

As I described in the article Tools of the trade, physics and theoretical physics, in particular, use Latex for scientific writing. This comes from a general tendency to prefer open-source and Linux-based tools. Moreover, Latex has the perfect equation typesetting. To write Latex, you can use whatever text editor you like, but I find Kile to be the easiest editor. Some people use Lyx or Overleaf.

Since the bibliography in a thesis is substantial, it is useful to use the proper instrument to cite it. I suggest using BibTeX since it is the most automatic and complete way to reference literature in Latex. You have to put BibTeX references in a separate .bib file, and cite it with \cite{...}. Figures and equations can be labelled with \label{...} and \ref{...}. Here is a short introduction to Latex by A. Cottrell, and a short tutorial on overleaf.com.

Presenting

When the thesis is done and delivered, you will have to present it (and sometimes defend it) in front of the examiners. This usually consists of a presentation that, in LU Physics, consists of 30 minutes or less. If your thesis needs to have a clear scientific message, this is doubly true for the presentation. In a presentation, everything needs to be purposefully presented with the objective of delivering a single, impactful, scientific message.

A good exercise is to think of your thesis and summarize the conclusion in 10 simple words or less. Now question everything: “Does this help me deliver this 10-word message?” Build your presentation on this.

Reason by blocks: the single presentation needs to build up to a single message; the single slide needs to have a single message that helps the presentation; the single figure and text needs to convey a single message that helps the slide. You get the jist.

If you have to revolutionize the structure you use in your thesis or cut out many results, so be it. A presentation has to be convincing and compelling, not a complete account of your work. In fact, quite the opposite. In the most prestigious conferences, you often have a few minutes to summarize years of work.

Also, in the presentation, the most important attribute is precision. Avoid touching on subjects you are unsure of and employ a specific and correct vocabulary adequate for your subject.

Following Up

It is fairly common that after the presentation, the examiners request some changes before agreeing on the final mark. Don’t be discouraged; scientific work and writing is a lifelong endeavor, and this is an excellent opportunity to polish your craft. Maybe your last opportunity to confront yourself with professionals in scientific writing.

If your work is particularly original and potentially impactful, your advisor can propose to publish it in a scientific journal. If that’s the case, you can use results, figures, and paragraphs you have produced in the thesis. You will discuss the type of article and the style to adopt with your supervisor.

In most cases, substantial revision is needed, because the format of an article is quite different from a thesis. A scientific article has a lower degree of self-sufficiency and a higher reliance on external sources. For example, in your thesis, you might need to define Hartree-Fock; in an article, it is not necessary in most cases since it is a well-known method and can be referenced. This might also imply that the notation you used might need a revision.

In this case, your supervisor will guide you very closely. It is good practice to offer a first draft, revised as asked. This first draft will probably need extensive correction, but again this is common. Having a publication out of a thesis depends on several factors not always under your control, but it certainly does feel good to have a test of the scientific maturity you have reached in such a short amount of time and will definitely help future PhD publications.

This concludes this guide. Don’t hesitate to contact me for more explanation and suggest modifications. Sorry if it’s long; I did not have time to make it shorter. To compensate, you deserve a Seal of approval to have arrived here!

*I had gifs on my website, but LU is stuffy and doesn't want my primo memes. Imagine a beatiful smiling seal*

Page Manager: peter.samuelsson@teorfys.lu.se | 2025-04-09