17

In academic writing, several articles on the topic of writing papers denounce the practice of "announcing" the topic.

As an example, if a paper were to read...

The goal of this study is to not be a study at all but to be a fictional paper of only one paragraph included as an example on stack exchange to demonstrate how a paper announcing its topic reads. This goal of this sentence is to fit some more content in with the previous paragraph on how sentences announcing their topic come across to the audience. In conclusion, this third sentence reinforces the point of the first sentence.

Why exactly is "announcing the topic" best avoided? What is a good way to explain it to others what "announcing the topic" does to a paper?

Examples of articles denouncing the practice include:

Agriculturist
  • 281
  • 3
  • 9
  • 1
    What point in the paper are you asking about? Starting a paper with "The goal of this study..." seems poor, but it can make a lot of sense to say something like that after a few paragraphs about the general field and previous work. "Widgets are important because of reasons. This pile of papers establishes the correct size of widgets. Other authors have also considered widget rigidity. In this paper, we study the surface texture of widgets." – David Richerby Nov 26 '19 at 10:31
  • 1
    @DavidRicherby I am an engineer copyediting a paper written by a non-native scientist. "this paper proposes... ", "the aim of this study is to...", "this paper builds.." are phrases I marked for removal. If this were a peer review I would not hold the paper for this, but simply make a comment to avoid announcing the topic. Since I am copyediting I do not want the work of this scientist to miss out on getting the attention it deserves simply because it reads like the example. I am looking to better understand the reasoning behind this guideline. – Agriculturist Nov 26 '19 at 20:37
  • 5
    @DavidRicherby Quite frankly, I rather like it when a paper just says e.g. "The contribution of this paper is to ..." – Azor Ahai -him- Nov 27 '19 at 14:38
  • 2
    @AzorAhai Exactly. You’re searching through tens of papers that might be relevant. You don’t want to have to read paragraphs and paragraphs just to see what a paper actually does. – David Richerby Nov 27 '19 at 15:42
  • 5
    This guidance is not uniformly applicable in all areas of academic writing. I was explicitly taught by my PhD thesis advisor, a highly published pure mathematician, to include sentences with phrases like, "This paper concerns..." The guidance linked from Arkansas State University illustrates with an example from a humanities field. In general, I think that scientific writing is much more likely to include such phrases. – benblumsmith Nov 27 '19 at 17:26
  • 1
    Tangengial: I remember an exchange with my advisor in which he wanted me to remove a somewhat metaphorical turn of phrase. "Ok, I'll try to make it more aseptic," I said. "Yes, aseptic!" he replied. "This is science." – benblumsmith Nov 27 '19 at 17:36
  • 1
    @Agriculturist I didn't know that one could also be born a scientist ... – Hagen von Eitzen Nov 27 '19 at 19:16
  • 1
    @HagenvonEitzen LOL... "non-native" in the context of my comment means a person grew up not speaking the native language. My native language comes so naturally, I am sometimes unaware of how to explain it. Native does not mean native-born in this context. This definition would imply the existence of a place, Scienceland perhaps, whose inhabitants called themselves Scientists. That can of worms is one I do not want to open so I am standing by my original usage of the term "non-native". – Agriculturist Nov 27 '19 at 22:19
  • 1
    „The goal of this study“ would be perfectly fine as it describes the fact it is about a study. – eckes Nov 28 '19 at 13:57

3 Answers3

32

In this answer, I am going to explain to you why you shouldn't announce what you are about to write anyway.

It is boring and redundant and a waste of real estate on the page.

Start with a claim, or a key observation. Those can be interesting. Don't talk about your paper in your paper, get to your paper! A sentence saying "The goal of this work is XYZ." can be eliminated without any loss of information. It has to be followed by an explanation of what the heck XYZ is, so beginning with that explanation is better. Fewer words, same quantity of information.

Although such papers are not sales tools, the psychology of writing advertisements does still apply, to academic papers or novels: Readers want to be hooked by the first sentence, interested by the first sentence, and that is going to serve you well, if they are interested in the opening they will be happy to read some less interesting sentences to gain some context and lead them into the discussion or story or article or advertisement.

In advertising, we say that on every sentence the reader is looking for a reason to stop reading and throw it away. The only reason they don't is because you have created a question in their mind, and they are reading to get an answer, or you are saying interesting things that they want to know. Don't give them a reason to give up.

That is less true for academic articles, but the advice is sound. Don't bore them from the first sentence. They probably know what the article is about from your article title and the rest of the context; the journal it was in, the keywords you selected, etc.

Amadeus
  • 101,174
  • 8
  • 132
  • 330
  • 15
    +1 for the opening sentence :) – Agriculturist Nov 25 '19 at 22:17
  • 11
    On the other hand, there is also this principle of 3-step content delivery: (1) Tell the audience what you're going to say, (2) say it; (3) then tell them what you've said. In academic writing this is accomplished by (1) Abstract, (2) Content, (3) Conclusion. – justhalf Nov 26 '19 at 06:03
  • 3
    This answer hits the nail on the head: “announcing the topic” is usually redundant. It’s worth noting a corollary, though: sometimes it’s not redundant, and in those cases it’s often not so bad. If there’s no way to get into the topic itself without some technical background, then a short orienting sentence before that technical stuff is very helpful. – PLL Nov 26 '19 at 09:40
  • 9
    I’ve had too many professors who jump right into formulas and solutions instead of stating what the question or problem is in the first place. – Michael Nov 26 '19 at 10:22
  • 1
    @Michael Speaking as a sometimes professor, they are doing it wrong. Read my advice: Start with something interesting, something motivating, that carries the reader through and makes them want to see more, or what's next. *Don't* start with something boring, or confusing, or redundant. This answer is about announcing. Presenting a puzzle is legit, asking a question is legit. Formulas and solutions are the answers to questions and at some time, somewhere a person invented those formulas in order to answer a question. Start with that question, that's interesting. – Amadeus Nov 26 '19 at 10:53
  • @PLL: An announcement is always redundant. "This paper is about XYZ." must be followed by another sentence explaining what XYZ is. If the audience already knows what XYZ is, then the following sentence gets more specific about XYZ, and we can rephrase that and start with it. An announcement is a crutch roughly 100% of the time, and can be eliminated. Like my opening sentence was designed to illustrate: Anybody reading my answer has presumably read the question and already knows what my answer is going to be about; I don't have to restate that. – Amadeus Nov 26 '19 at 11:11
  • @justhalf That principle doesn't apply everywhere. I wouldn't write a story with a full story synopsis up front, or any story. And in academic papers there is often a page limit, so sticking to triple-redundancy is not always possible or useful. I publish academic papers, I think of it as more (1) Problem overview, (2) Mechanical Details, (3) Resolution and Implications. – Amadeus Nov 26 '19 at 11:22
  • 10
    You write quite well, but honestly, in academic papers my first goal when reading a paper is just sorting out if the paper is relevant or not. And in my experience, papers are more likely to be irrelevant than otherwise. Creating mystery might be tolerable when the writing is good and easy to understand, but in more complex matters it makes the content more obscure than needed. Furthermore, I'll get really pissed if you hook me up reading something I never should have read in the first place. – Mefitico Nov 26 '19 at 13:11
  • 1
    @Mefitico I write and review academic papers (6 this year), it is part of my job. I think the context of the Journal (or Conference) in which it is published, your Paper Title, and your keywords is sufficient to tell whether the paper is relevant or not, or definitely should be if the author is doing that part of their job correctly. Following that, the Abstract should indeed be an abstract, but if your journal is Statistics and your title is "Robust fitting of the XYZ distribution", isn't it redundant to say "In this paper we will demonstrate robust fitting of the XYZ distribution"? – Amadeus Nov 26 '19 at 14:42
  • 1
    @Amadeus-Reinstate-Monica When I'm reading "for research purposes" or if I've already seen the conference presentation for that paper, then indeed, title and keywords suffice. But when I'm implementing something for work, if a paper claims to "robust fit of XYZ distribution", then I care if it presents the algorithm or just cites it, if it takes too long to run, if there is any major simplification, if the quality of results is acceptable for my application and so on. I often find papers that exaggerate claims on abstracts, obscure relevant simplifying hypotheses, reach poor results and so on. – Mefitico Nov 26 '19 at 16:59
  • 1
    Case in point, if the title of the paper is "Novel Algorithm with Guaranteed Robust Fitting of XYZ in polynomial time, demonstrated in C language with optimal accuracy", I know I've found what I"m looking for. If the title is "Robust fitting of the XYZ distribution" but could have been "Monte-Carlo Approach for Finding Best Possible Accuracy for Fitting XYZ Distribution, Using the Simplest and Most Well Known Fitting Algorithm that Mefitico Knows won't Solve his Problem". Then it's a waste of my time to read it, and no good writing or scientific merit of the paper will make it worth it. – Mefitico Nov 26 '19 at 17:08
  • 1
    @Mefitico Then you should approve of no restatement of the Title in the abstract, it leaves more words to provide details about what was actually done. Which is the question I asked you; if the first sentence of the Abstract reflects the Title, an announcement about what the paper is about, then isn't it redundant? Wouldn't it be better to use those words to mention something important about the content of the experiments, or method, or accuracy, or overhead? The topic here is whether announcements make sense, it isn't about "how to write a good Abstract" (or good Title) in general. – Amadeus Nov 26 '19 at 17:10
12

There are reasons to repeat yourself, but they differ to what you imply.

  1. Academic readers are usually skimming through hundreds of papers to find the results relevant to their current work, so

    1. Try to tell them everything they need to know in the title

    2. If that fails, try to tell them everything they need to know in the abstract, including results and conclusions. (Some journals discourage this, IMO wrongly).

    3. If that fails, still be careful not to withhold information early on. i.e. in the introduction don't say "We test to see whether X predicts Y", say "We demonstrate that X predicts Y (R2=0.8, p=0.001 - or indeed better stats if your audience will understand them).

  2. In a large paper on a complex topic it's easy to get lost, so do include signposting to help the reader remember their context. E.g. "The link between X and Z is a relevant consideration for our study of X and Y, therefore in the following section we discuss existing literature discussing links between X and Z."

But don't use the rhetorical device of "tell someone what you're going to tell them, tell them the thing, tell them what you told them" for the sake of driving a point home, as you would in rhetoric, or delivering spoken material as a teacher, etc. (Excepting the first sentence of the discussion/conclusions section where it is usual to summarize what you already told them, in shorter form). You repeat to make it easy for readers to find the appropriate part and remember context.

Sideshow Bob
  • 221
  • 1
  • 3
  • Comment 1/2 : This is the exact argument I am up against. Go to this article: https://www.americanscientist.org/blog/the-long-view/the-science-of-scientific-and scroll to reader expectation for the structure of prose. If most text reads like the gibberish in the first example then I would agree that repeating is helpful. If the text reads like the second where everything is where the reader expects it then it seems like more reader energy is consumed than necessary. – Agriculturist Nov 27 '19 at 19:49
  • Comment 2/2: The whole point of the methods section is to avoid repeating the experimental details necessary for replication when discussing results. If done correctly the results section should be less complex. .. The same is true of every other section using a standard scientific format. – Agriculturist Nov 27 '19 at 20:10
1

If you think of yourself as reading a novel, rather than a scientific paper, it becomes very clear. If a novel started with 'In this novel, you will read..' you might find it awkward, or at least old-fashioned - I think there are Victorian novels which take this approach, but it is not the way of modern fiction. It removes the reader from being an active participant 'in the moment' to being a third-party observer, and it injects the voice of the author very load and clear.

In a scientific paper, you want the science to speak for itself, and as a reader, you don't want to be concerned with the authors in particular.

Another key message from novel-writing is to show people something, rather than telling them. This can also apply to scientific writing, and is relevant here I think.

Jay Moore
  • 157
  • 3
  • 2
    I really like the point about "the science should speak for itself." The whole point of experimentation is to get science to tell us its story. That story is what we are trying to tell the world when writing a scientific paper. If that story is muddied with the story of the author "announcing" the topic, or if it is already so muddy that the author feels compelled to "annouce" the topic then the reader may question if the story of the science is worth reading. – Agriculturist Nov 26 '19 at 19:41
  • 4
    Personally, when I pick up a novel, I like to look at the back of the book and read "In this novel, you will read..." It helps me decide if I really want to read it. Books that have only testimonials in the back: "This book is awesome, you must read!" really demotivate me and I tend to not read them, since I couldn't get more information that would help me make the final decision of reading it. But of course, once I decided to read it, I don't want the novel itself to start that way... :) – msb Nov 27 '19 at 04:22
  • 3
    @Agriculturist You can't always rely on the science to speak for itself. A key part of scientific publishing is novelty, and explicitly pointing out how your paper differs from what's already in the literature is one way to demonstrate novelty. As I mentioned in my comment to the question, "In this paper we..." is a perfectly reasonable way of contrasting your work with a description of what has gone before. Sure, you can argue that the words "In this paper" are redundant but, jeez, it's only three little words in ten plus pages of text. – David Richerby Nov 27 '19 at 10:06
  • 6
    Scientific papers are not novels, and they are fundamentally different from novels. For example, a key element in many novels is suspense, which is exactly what you don't want in a scientific paper. In a whodunnit, the last sentence is "Col. Mustard killed him with the lead pipe in the conservatory"; in a scientific paper, that would be the title! A scientific paper is exactly there to tell. The reader doesn't have time to do all your data analysis again: you need to tell them what the conclusions of your research are. You need to tell them why they should read your paper. – David Richerby Nov 27 '19 at 10:10
  • 1
    Scientific papers are also fundamentally the same as novels, in that their impact derives from enticing people to read them. Some people might not mind papers that start with 'In this paper...', but enough readers will be put off (N>0) to make it worth not doing that. – Jay Moore Nov 27 '19 at 14:02
  • 4
    @JonathanMoore What?!? I’m sorry but I can’t even remotely imagine somebody taking so much offence at the words “In this paper” that they refuse to read a paper that is relevant to their own work. We all have to read so many papers in broken English (or, from the other side of the coin, read so many papers in English when that is not our native language) that, honestly, a paper whose only sin is to say “in this paper” is an unimaginable luxury. Where can I find such a gem? – David Richerby Nov 27 '19 at 15:41