“Patchwriting” is a Wegmanesque abomination but maybe there’s something similar that could be helpful?

Reading Thomas Basbøll’s blog I came across a concept I’d not previously heard about, “patchwriting,” which is defined as “copying from a source text and deleting some words, altering grammatical structures, or plugging in one synonym for another.” (See here for further discussion.)

As Basbøll writes, this is simply a variant of plagiarism, indeed it’s an excellent description of what some of the craftier plagiarists actually do. I’m reminded of the statement of history professor Matthew Whitaker’s publisher that Whitaker couldn’t have plagiarized his recent book because he’d assured her that he’d run it thru two different plagiarism programs, or something like that.

As the saying goes, if you have to run your book through two plagiarism programs, you’re already in trouble.

OK, so I’m 100% with Basbøll that “patchwriting” is plagiarism. And, like Basbøll, I’m a bit disturbed if some people think that patchwriting is “virtually inevitable as writers learn to produce texts within a new discourse community.”

But I wonder if there’s something similar to patchwriting that could serve the same function but much more constructively. Check out this paragraph that Basbøll quotes from Pecorari, “Academic Writing and Plagiarism” (1999):

In a study of the course of the progress of second-language writer through a business course, [P. Currie] found that the student, Diana, worked diligently in the early weeks of the course to raise the level of her writing assignments, but was at real risk of not receiving the grade she needed to stay in her program. Eventually Diana hit upon the strategy of repeating words and phrases from her sources; in other words, she began to patchwrite. From then on her teacher’s feedback was more positive.

My take on this is that, if student Diana did this right, it could be a great way for her to transition to learning to write on her own.

Just for example, imagine how the above quoted paragraph could be “patched”:

Pecorari (1999) writes of “a study of the course of the progress of second-language writer through a business course.” In this class, Diana (the student in the class) worked hard (Pecorari uses “diligently”) to raise the level of her writing assignments write better. Diana was worried about not getting a good grade. Then Diana did patchwriting: she did “repeating words and phrases from her sources.” Her teacher liked that. Her teacher gave Diana positive feedback.

I’ve purposely written this in one take, in a somewhat awkward style to imitate how a student might do it. I also put in a strikethrough to illustrate another way that a student might paraphrase but in an honest way. And, even so, it’s not perfect; one might say it still teeters on the edge of plagiarism, even with the sourcing, because I (playing the role of the hypothetical student) am adding nothing—like Ed Wegman or Frank Fischer or Matthew Whitaker (but in a more honest way), I’m merely regurgitating.

A better approach could be a complete blockquote followed by a summary and a reaction. For example, suppose the (hypothetical) student paper went like this:

In her 1999 book, Pecorari writes:

In a study of the course of the progress of second-language writer through a business course, [P. Currie] found that the student, Diana, worked diligently in the early weeks of the course to raise the level of her writing assignments, but was at real risk of not receiving the grade she needed to stay in her program. Eventually Diana hit upon the strategy of repeating words and phrases from her sources; in other words, she began to patchwrite. From then on her teacher’s feedback was more positive.

In my own words: Diana repeated words and phrases and her teacher liked it. Diana got better grades.

My reaction: Pecorari thinks it worked. I like the idea too. I want to learn to write on my own. Will patchworking work for me?

That’s probably not a good imitation on my part of student writing. But I have two points here:

1. Patchwriting, even with full sourcing, is empty as a means of self-expression but could still provide some useful practice. Just as we can learn from looking up classic chess games and playing them out on the board, maybe novice students could learn by re-expressing source material.

2. Shuffling around the words of others seems like a bit of a dead end, so I think its limitations should be kept in mind.


Unlike Basbøll and (I assume) Pecorari, I have very limited experience as a writing teacher, so these are just some quick ideas I’m offering up. Conditional on these caveats, here are my thoughts:

– Pure patchwriting as in the definition at the top of this page seems like a terrible idea.

– But I could see it making sense to encourage “patching” (as in my example above) and block-quoting-and-explaining (as in my other example) as a way to learn.

– We may need to somewhat separate the goal of learning to put words and sentences together, and the goal of expressing oneself. A writer needs to learn both these skills.

35 thoughts on ““Patchwriting” is a Wegmanesque abomination but maybe there’s something similar that could be helpful?

  1. The fundamental error here is to think that the literary problem of saying what another text says without plagiarising it exists. There is the problem of representing another text, i.e., saying that a given text says something. This is managed easily by quotation and paraphrase. Then there’s the problem of saying what you think. And you just go right ahead and do that. No one should ever face the problem of doing in there own writing what someone else has done in theirs.

  2. How does this work for writing code for stats? I’m not thinking complex code that might be published, but if someone wanted to make a plot with certain characteristics, they might find an R gallery and choose a few plots with the right characteristics (or just do ?hist or example(hist) to get examples), take bits of the code, but change the variable names. So I see

    hist(x, freq = FALSE, ylim = c(0, 0.2))
    curve(dchisq(x, df = 4), col = 2, lty = 2, lwd = 2, add = TRUE)

    and patchwrite it to

    hist(myx, freq = FALSE, ylim = c(0, .25),las=1)
    curve(dchisq(x, df = 4), lty = 2, lwd = 2, add = TRUE)

    I assume most people look at other people’s code to help when they are uncertain of a function’s syntax. This seems a good strategy for learning, which is this blog’s point for learning English, but it is patchwriting. Are similar issues relevant here to patchwriting the text?

    • I think the formal structure of code leads to a lot less flexibility in how it’s written and what can be done than in natural language. Patchwriting code because there’s really only a fairly narrow set of ways to use “hist” makes sense. But, even parsing English is nontrivial, and there are often multiple possible meanings to any given sentence. “Time flies like an arrow, fruit flies like a banana” and all that. I guess my point is that patchwriting English is problematic precisely because natural language is so expressive that any given thing that is written expresses a fairly specific viewpoint: that of the original author, and rewriting another person’s viewpoint fails to qualify as an independent viewpoint.

      I think Andrew’s point is that sometimes people have to learn to put words together before they learn to put ideas together. Perhaps patchwriting helps with learning to form clear sentences even if it fails to teach anything about ideas or points of view.

  3. For the patchworking as chess practice analogy to fit, chess students would need look up classic chess games and then play the classic moves in a *different but similar* order, which doesn’t seem to be as helpful. I think a better use of patchworking-as-pedagogical-practice is for students to patchwork a poorly-written passage into respectability.

    I do like the classic chess games analogy, but for that analogy to fit, it would be better for students to study the choices that a [grandmaster/great writer] has made: for example, Percorari wrote “From then on her teacher’s feedback was more positive” but could instead have written “Her teacher’s feedback was more positive after that” or “Subsequent feedback from Diana’s teacher was more positive.” I think it’s a good idea for students to think about how those sentences have different values, both as individual sentences and in context of a larger passage.

  4. Hunter Thompson once retyped “The Great Gatsby” and “A Farewell to Arms” in their entirety to know how it feels to write a great novel. He was also learning about the writing style. Many authors claim that they learn to write by copying (often privately, so not really plagiarism). Some authors say that translating works is a way to learn to write. Same for painters, they start copying the masters.

    • The privacy of such imitation is crucial. As far as I know, none of Thompson’s published work contains the results of those typing experiments. (I agree that it was a good way to find out how those novels “work”. Just as imitating the style of a master painter can teach you something.) I’ve seen people pass off their translations of Kant and Hegel as paraphrase, even interpretation. E.g., “Hegel argued that…” after which you get three paragraphs of translated Hegel rather than three paragraphs of the author’s understanding of Hegel. That’s plagiarism too.

      • Hmm.. that’s a little sketchy as to whether it constitutes plagiarism. The author is giving Hegel credit for the ideas after all, but is not putting quotes around it, because it’s not Hegel’s words… it’s translated. Probably what’s needed is a typographical tool that isn’t available, a kind of “trans-quote-mark” in which the contents are understood to be a translated quotation of the original.

        • It’s straight plagiarism. To see this, imagine the case of translating it back into German. Either the result will be a plagiary of Hegel (let’s agree that if I write, “Pecorari says that in a study of the course of the progress of second-language writer through…”, and then use her words for three of what look like my paragraphs of paraphrase, I’m plagiarising) or the result will be what looks like a German paraphrase of Hegel. But who did the paraphrasing? The author, or the the translator? In fact, the supposed “paraphrase” only came about out of translation errors. On this norm, it would be possible to produce a paraphrase of any author by passing it through Google Translate a couple of times and then cleaning up the result for grammar.

        • But if the writer says

          “In so and so such and such an author says ‘foo bar baz quux'”

          where ‘foo bar baz quux’ is a translated version of what the author actually says… then it’s lying. The author said no such thing. I suppose the way to deal with this is to write “‘oofay arbay azbay uuxqua’ (foo bar baz quux)” where the parenthesized words are the translation and the quoted words the original… this would be easier to understand if I were using real language.

        • There’s nothing dishonest about saying a sentence like “In the Phenomenology of the Spirit, Hegel talks about the ‘the forms of the unreal consciousness'”. In one’s referencing one should simply make clear whether one is using a translation of Hegel (from which one should then quote accurately) or the German original (in which case one simply specified that “all translations of Hegel are my own”). The translations should in both cases be in quotation marks, and only in some cases will it be necessary to quote the original German in brackets (as you suggest).

          What is dishonest is to write, “In the introduction to his Phenomenlogy of the Spirit, Hegel argues that the completeness of the forms of the unreal consciousness will arise from the necessity of their relationship and development itself.” That’s my quick touch-up of the Google translated version. An actual English translation gives us: “The completeness of the forms of unreal consciousness will be brought about precisely through the necessity of the advance and the necessity of their connection with one another.” That is, if all you’ve done is translated a sentence (no matter how well) then you cannot present it as your own interpretation of Hegel’s argument.

  5. I always found this difficult when working with my students, because the whole tertiary education system (at least in Australia for undergrads) is really based on how well you do this. It’s lovely to think we are teaching students to think for themselves and add their own thoughts to their work, but in reality that’s how to get bad marks in the current university system. The way to get great marks is to write exactly what the lecturer has said in their notes and throw in some ‘paraphrased’ ideas from references on the reading list. I learned it during my undergrad, and my students now get amazing marks when they follow that advice (and no, I’m not the marker).

    • The thing is, the “real” world often works the same way — lots of people have jobs where their main responsibilities are to tell people what they want to hear and sound smart while doing it.

  6. Starting with a template that goes “paraphrased cited bit”, then “this suggests that [students idea or reaction]” would be a pretty good building block, in my opinion.

  7. Interesting. So everyone here seems to agree that patchwriting is wrong when it comes to Wegman. However when a liberal author adopts the same practice, it is “within the normal boundaries of research methods” and “an entirely fake controversy ginned up wholly by wingnuts”. (http://publiceditor.blogs.nytimes.com/2014/08/12/was-an-accusation-of-plagiarism-really-a-political-attack/)

    Maybe I am being too skeptical, but the view of the acceptability of the practice seems to depend on whether a liberal or conservative meme is being followed. To be clear, I am not accusing Prof Gelman of a double standard – AFAIK he has never commented on the Perlstein controversy. However, I do find it interesting how every academic authority consulted by the Times strongly defended the use of patchwriting by a liberal author when attacked by a conservative. I wonder if they would so quick to defend Wegman.

  8. finat might want to study the Wegman case a little more.

    1) The Wegman Report was ~1/3 plagiarized, and sometimes distorted, as part of an effort to mislead Congress.
    2) Said, Wegman, et al had a paper retracted for plagiarism.
    2) Plagiarism was found in several of Wegman’s students’ dissertations.
    3) Wegman and Said lost editorships at WIREs:CS for mosaic plagiarism, some with ludicrous errors, in papers for their own journal.
    4) And lots more.


    • Also there was that time they stole from Wikipedia and garbled the math in the process. That was pretty embarrassing, especially given that he’d received a lifetime achievement award from the American Statistical Association.

      • Speaking of Wikipedia: The article on plagiarism on Wikipedia is delightfully frank & pragmatic when it says:

        “The idea of plagiarism remains problematic with unclear definitions and unclear rules. The modern concept of plagiarism as immoral and originality as an ideal emerged in Europe only in the 18th century, particularly with the Romantic movement. Plagiarism is not a crime per se….. the concept does not exist in a legal sense. “Plagiarism” is not mentioned in any current statute, either criminal or civil…..The concept of “self-plagiarism” has been challenged as being self-contradictory, an oxymoron,and on other grounds.”

  9. People are getting so nitpicky, it might just make sense adding a note to your work: “I claim none of the above work as original nor mine. Lines or entire sections may have been lifted verbatim from other sources.”

    “representation of another’s work as one’s own original work” remains an essential element of the crime of plagiarism right?

    If I just outright abjure ownership, I solidly defend myself against any potential claims of plagiarism. And save myself the headache of painstakingly delicate citation of every damn thought & idea.

    • Rahul:

      I think you’re missing the point here. My problem with plagiarized work is not about ownership or originality, it’s about not supplying the source, which can greatly interfere with the information-passing that is a key aspect of writing.

      Basbøll and I discuss this in our two papers, here and here.

      Abjuring ownership does not defend yourself against claims of plagiarism. The point is not that sections “may have been lifted,” it’s for the reader to know where they have come from.

      • Andrew:

        In which case, your conception of what constitutes plagiarism seems to be in conflict with other authoritative sources. The misrepresentation of ownership or originality seems integral & essential to most definitions of plagiarism that I dug out. Ergo, if you deny ownership I suspect you cannot be held guilty of plagiarism.

        I’d love to hear other opinions or interpretations of this.

        e.g. (1) Wikipedia: wrongful appropriation of another author’s language thoughts, ideas, or expressions and the representation of them as one’s own original work.

        (2) UNC Chapel Hill: the deliberate or reckless representation of another’s words, thoughts, or ideas as one’s own

        (3) UNSW-Australia: using the words or ideas of others and passing them off as your own.

        (4) MIT: Plagiarism occurs when you use another’s words, ideas, etc. and do not acknowledge that you have done so.

        (5) Columbia: plagiarism has been defined as the wrongful act of taking the product of another person’s mind and presenting it as one ’s own

        • Rahul:

          Hmm, maybe you’re right about this one. From my perspective, the problem when someone misleadingly doesn’t give the source of a quotation or an idea, and problems arise when the source is obscured (for example, Wegman messing up on the math with the wikipedia copying, an error that I think might not have occurred had he been more open that he was taking it from the Wik). The disclaimer you say is fine, in some way, but it would make any article essentially useless to me.

        • Andrew:

          I think it’d still be pretty useful. Say, I took your radon sampling paper & removed all citations. Would it suddenly become totally “useless” to a reader? I don’t think so. To someone like me, it’d be perhaps 80% as useful as the original.

          OTOH, I’m not even saying get rid of citations. Just add the disclaimer. That way you are insulated against any odd, accidental plagiarism & you can be just more relaxed in general about citations. And people cannot nitpick about “patchwriting”, self-plagiarism and other such gray areas.

          I can cite the really important stuff. On a lazy day I can skip some citations or even add verbatim text where I think the original author did a good job as it is in stating what needs to be stated.

        • Rahul:

          Maybe my paper is ok because you trust me. But for other readers it would be good to know where the data and methods came from. Finally, I have no problem adding verbatim text from other sources; just put it in quotes. Just like how I do it on the blog. “As Smith says, ‘….'” Works just fine.

        • I think what you wrote brings out another very crucial point: Do I trust a paper based on the comprehensiveness of its citations more? Or based on who wrote it, what Journal, what university, have I read that guy before etc.

        • Rahul:

          A lot of people trusted Karl Weick, who teaches at a top university, is well respected, etc. But in his plagiarisms he distorted his source, destroying much of its value in a social science sense (see my papers with Basbøll for discussion of this case). Had he been open about his sources, much of this could’ve been avoided. It’s not an issue of “painstakingly delicate citations,” he just should’ve said he was quoting a damn poem and given the source.

          In many ways, his articles become less readable and less comprehensible due to the constraints required in burying his sources.

          And this happens a lot. People want to get credit for things they didn’t write, so they go to extra effort to hide the sourcing, and the result can be a garbled mess.

  10. Rahul’s suggestion is useful, if the disclaimer is at front of paper: I could stop reading right there.
    It’s like seeing comments from”Anonymous.”

    The Net is filled with poorly-sourced junk one need not read, and it would indeed be helpful to know which papers could quickly be ignored.

    In the real world, one cannot go spend infinite time crosschecking, so people accumulate credibility assessments.
    A: If someone builds a record of sloppy scholarship and bring wrong again and again, people start ignoring them.
    B: At the other extreme, people can build a record if bring careful, being right often and fixing mistakes … And they get taken seriously … even when making surprising claims.

    • Would you stop reading even if the name was someone whose work you trusted & respected?

      It isn’t the same as “Anonymous”. You have certainty on who wrote the stuff.

    • John’s response to Rahul is the right one. The only usefulness of the disclaimer is that it flags a paper that is useless for any scholarly purpose and only informative to someone who is ready to swallow it whole. It basically says, “I don’t know what I’m talking about and don’t care.” In some (overly) formal sense it may not constitute plagiarism, but it does in the following sense: If I “write” a YA novel by copying passages from Hunger Games, the Maze Runner, and Divergent, changing the names and places and a couple of details to make it a consistent narrative and then add Rahul’s disclaimer to the front matter I’d still get sued. We can imagine producing a low-cost textbook by a similar procedure. A serious publisher, like any serious reader, will simply not bother with the text.

Leave a Reply

Your email address will not be published.