In scientific papers (in my case it's usually in computer science) it seems to me that some techniques which help in explaining concepts and technology are not used that much.

For example a metaphor often helps one to understand abstract concepts. Not the most useful example but something like "Monads are like burritos, they are not plain, but something wrapped in something else".

Another such technique would be explaining a specific (and simple) case of something - and maybe a second and third case - only to reveal later on what they are actually specific cases of. By successfully following the simple examples and seeing what they have in common, it is then easier to get an intuition about what the abstraction is all about. As a bonus, it might also make studying the material more enjoyable because it just seems easier.

Now, I don't remember reading stuff like that all too often in scientific papers, but I often find them in online tutorials, blogs and such.

Is their use generally discouraged, and if so why?

I might be wrong in at least two ways:

  1. These "tricks" are used. I just don't see them.
  2. They have a disadvantage I did not think of. I just think they can be very beneficial, and the extra time spent reading is often worth it.

I think the online book Learn You a Haskell for Great Good! does most of the things I mean. It also includes some jokes and funny drawings which are not what I'm talking about.

  • I can't say if such techniques are discouraged, but I know I prefer reading science writing which is clear and accessible. Metaphors and other literary devices often help with readability. Sharon Begley sharonlbegley.com is one of my favorite science writers specifically because she makes everything so clear and easy to understand. Her writing is lively and colorful — like a cheery blog post rather than a dry technication recitation. Feb 6, 2017 at 16:16
  • Metaphors, if they work, can become normalised. Many many years ago, I was explaining to colleagues that choosing from a list of options in a system I'd written for a character-based terminal was just like picking things from a menu: now a common word in user interfaces. No, I didn't invent the current usage but obviously someone had the same idea for a metaphor :)
    – Tony Linde
    Feb 6, 2017 at 17:24
  • A small clarification: Please don't get hung up on metaphors. And I never meant dumbing anything down or saying things which are factually wrong. If you deny that a lot of scientific literature could be written in a way that is easier to understand while being just as correct, don't expect me to listen to you. Often that would mean just adding something.
    – Higemaru
    Feb 11, 2017 at 14:20

3 Answers 3


There are two aspects to writing style: there is what does the best job of explaining a concept, and there are the shibboleths that determine if a certain group is going to accept the document. Unfortunately, when submitting a document for publication, you have to consider both how stylistic decisions affect comprehension and what shibboleths affect publishing decisions.

The use of analogies to explain concepts is great for comprehension. Their use is not a "trick". It is standard practice for effective exposition. All good popular science that I have seen uses analogies for effective exposition.

However, the academic profession is plagued with shibboleths. One very common shibboleths is that works should not be "popular" in orientation or style. They are supposed to talk only to the in crowd. Being incomprehensible to the general public is a sign that you are a member of the academic club, that you have paid your dues.

Might the use of analogies to explain concepts make your paper sound too "popular" for an academic journal? Maybe. Such prejudices are not universal, nor are they consistent. As with any other publication, it pays to spend some time to get a feel for the tone of the publication you want to submit too. If other papers in that publication use the same devices, you should be fine.

Another reason that you may find few uses of analogy in scientific paper is a cognitive bias called "the curse of knowledge". The curse of knowledge is a bias which makes it difficult for us to understand how other people could possible not understand a concept once we understand it ourselves. Even if we learn a concept ourselves by way of analogies, once it has clicked in our brains, we recall it by its formal name alone, and it ceases to occur to use that anyone might need the same analogy we learned from in order to understand the point. Indeed, the analogy can now seem an unnecessary circumlocution that simply slows down explanation.

Someone suffering from the curse of knowledge is unlikely to turn to an analogy to explain something because it is unlikely that it will occur to them that it needs to be explained. The lack of analogies in many scientific papers may therefore be due purely to the curse of knowledge.

The good popular science writers may simply be those who still recognize when a concept needs to be explained (even when they have internalized it themselves) and can come up with an appropriate analogy to explain it.

Unless you have to suppress you analogies to pass the shibboleths of the journal you are submitting to, therefore, the apt use of analogies to explain key concepts in your paper is a very good thing.

  • Thanks! You magically inferred that I meant working analogies, not abstruse ones. Seems I suffer from the curse of knowledge myself ;-)
    – Higemaru
    Feb 6, 2017 at 19:44

Academic publications have their own individual styles and the ones that I have read (few it must be admitted, when it comes to computer science) generally discourage the use of things like metaphors and other figures of speech, such as hyperbole, similes and understatement.

One reason for this is, I think, is that these things don't cross cultures and language translation very well. For example, I don't what a burrito is. Someone who is a native speaker of Japanese can have more trouble understanding figurative language written in English than someone from say America.

Having said that, popular science journals, even quite technical ones do include articles that use these techniques because they can be very useful for communicating clearly. Figurative language, used well, can make ideas and concepts clearer. For example, I still remember an article I read in the 1980s about computer processors and the whole thing was framed in term of magic.

  • 1
    Hit the nail on the head. While academia does have its own kind of language, adhering to seemingly arbitrary rules such as the use of passive voice, it's not just snobbery. For some people, the metaphor will only confuse matters. It's also not very professional; I've read silly analogies and metaphors in physics papers and it invariably induces a dramatic groan and eye-roll combination. Feb 6, 2017 at 16:52
  • I'm surprised you don't know burritos. A bad metaphor (and I would include those which don't translate) is of course useless. But surely there are some universally understood concepts. In a scientific paper, you can also expect some fundamental ideas in that science to be known to a reader of such a paper.
    – Higemaru
    Feb 6, 2017 at 19:08
  • 1
    That's the thing: you are surprised I don't know what a burrito is. But why should I? I have never eaten one. I have lived in two different countries and never been offered one. In fact, I have visited several countries and don't remember seeing one on a menu. Cultural references don't translate easily. Scientific language is much more universal. Of course you could expect readers of a certain level to understand fundemental scientific ideas: you can't assume they comprehend cultural references, idiom or figures of speech. Feb 6, 2017 at 19:45
  • Forget about the burrito. I also did not have cultural references or figures of speech in mind, I can't even think of a usable example using them. Anyway, is there anything you can guarantee to be universally understandable?
    – Higemaru
    Feb 7, 2017 at 12:49
  • 1. Formal, technical language is understood by a large percentage of the audience that would read a particular text, even if it isn't understood by the general public. 2. Metaphors can depend on cultural knowledge -- that is all I meant by 'cultural references'. Feb 7, 2017 at 17:00


The one fundamental rule in scientific writing is clarity. Analogies, metaphors, and other figures of speech are never clear in their meaning but open to interpretation.

When I say, for example, that a cell in biology is "like a power plant that burns fuel to gain energy" then that analogy will give you a completely wrong picture of what goes on in a cell. Some metaphors and analogies have even stifled scientific progress for decades, such as the view of the human psyche as a steam engine (Freud) or as a computer.


For that reason, reputable scientists avoid the use of analogies and metaphors and reputable journals discourage them.

Science instead uses its scientific terminology and requires the reader to make the effort and work for their understanding. Science is not for the mentally lazy.

Figures of speech are closely related to clichés. Like metaphors and similes, figures of speech provide a writer with a colourful or forceful means to draw attention to a particular point but should be avoided in academic writing. (from What to avoid in formal writing)

  • 1
    Here speaks the voice of the shibboleth.
    – user16226
    Feb 6, 2017 at 18:20
  • I just can't agree with you on this. It is exactly this elitist (sorry) attitude that I don't want to be dragged into. Is anything not open to interpretation? Why would metaphors never be? I also tend to see a lot of mental laziness in not caring to explain things properly. I have yet to encounter anything that is actually hard to understand even when there are no hindrances like missing information. That's probably another symptom of the curse of knowledge.
    – Higemaru
    Feb 6, 2017 at 19:29
  • 1
    There is absolutely nothing elitist about this. Academic journals must provide precise, universally understood descriptions. Metaphors are anything but precise. If I read a photonics article that was riddled with Chinese proverbs, colloquialisms and metaphors then I'm sure that it would make the author's work very hard for me to understand. Feb 7, 2017 at 10:08
  • I'm not advocating replacing anything with metaphors. It's not even so much about the metaphors, sometimes just stating what something is not can help tremendously.
    – Higemaru
    Feb 7, 2017 at 12:39
  • 1
    When I explain optical coherence tomography to my family I say that it is like ultrasound, but with light instead of sound. Except that it's not. Not at all. I would never use this comparison in a paper because it's factually wrong. It does, however, give the layman a rough idea of the technology using context that they already know. Feb 8, 2017 at 16:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.