Careful scholarship is supposed to protect us from chains of whispers, where texts get distorted via paraphrase and summary so that secondary and tertiary works fail to accurately convey the sense of the original. But it only achieves this goal if scholars are prepared to check their sources all the way back to primary texts at least once in a while. If nobody bothers to do this then a mistake can persist in the literature for decades. If everyone looks only at secondary and tertiary sources, a field is vulnerable to citogenesis.
An example of this effect is the claim that a decimal-point error led scientists to believe that spinach was a good source of iron. This was investigated by Ole Bjørn Rekdal, who wrote:
Many of the messages presented in respectable scientific publications are, in fact, based on various forms of rumors. Some of these rumors appear so frequently, and in such complex, colorful, and entertaining ways that we can think of them as academic urban legends. The explanation for this phenomenon is usually that authors have lazily, sloppily, or fraudulently employed sources, and peer reviewers and editors have not discovered these weaknesses in the manuscripts during evaluation. To illustrate this phenomenon, I draw upon a remarkable case in which a decimal point error appears to have misled millions into believing that spinach is a good nutritional source.
Ole Bjørn Rekdal (2014). ‘Academic urban legends’. Social Studies of Science 44:4, p. 638.
The point that Rekdal makes is that it is misleading to cite a source, if you know or suspect that the source does not provide an adequate justification for the claim. A scholar tracing the spinach claim, Rekdal says, might find their way to Hamblin (1981), which tells the story without citing any sources for it, and be tempted to conceal this inadequacy:
In such a case, it may be tempting to suppress what I just discovered, that the fascinating decimal point error is thus far an undocumented piece of information, what we usually call a rumor. I could correct my first statement, for example, in this way:
The myth that spinach is a good source of iron has its origin in a decimal point error in the 1890s. German scientists discovered the error about 40 years later (Hamblin, 1981: 1671).
Here, I have done my academic duty by consulting Hamblin directly, and I have corrected my statement in a manner that corresponds much better with what Hamblin actually wrote. If I publish this text, I am still guilty of spreading a rumor, and also of disguising its rather obscure origin through a reference to a prestigious scientific journal. I very well know that the decimal point error story suffers from a lack of documentation, but I cannot resist the temptation to use it, perhaps because it fits so nicely into the argument of an article I am writing.
Rekdal, pp. 643–644.
In such a case, it is important to either check the primary source yourself, or if you cannot, to indicate to the reader the inadequacy of the secondary source you consulted. The rise of publicly accessible online searchable corpora makes it possible for anyone to investigate these cases, and if you do then you will find that scholars have passed on a lot of rumours.
A couple of examples that I looked at which turned out to be rumours were the claim that James Joyce condemned Hamlet as a “failure” and the claim that T. S. Eliot plagiarized ‘The Waste Land’. In this article I’ll look at another instance of the phenomenon, in which Michael Dummett’s imperfect recollections of the details of a science-fiction story have been copied for decades by philosophers without anyone checking against the original.
If there are causal chains running in the reverse as well as in the usual direction, there is a possibility of causal loops. They often occur in science fiction: one such story that amused me concerned a fifth-rate but conceited artist. One day he is visited by an art critic from a century ahead, who explains that he has been selected for time travel so that he could interview the artist, who is regarded, in the critic’s time, as by far the greatest artist of the twentieth century. When the artist proudly produces his paintings for inspection, the critic’s face falls, and he says, in an embarrassed manner, that the artist cannot yet have struck the inspired vein in which he painted his (subsequently) celebrated masterpieces, and produces a portfolio of reproductions that he has brought with him. The critic has to leave, being permitted, for some unstated reason, only to remain for a limited length of time in the past, and the artist manages to conceal the portfolio, so that the critic has to leave without it. The artist then spends the rest of his life producing the originals of the reproductions by carefully copying them in paint.
Michael Dummett (1986). ‘Causal Loops’. In Raymond Flood & Michael Lockwood, eds. The Nature of Time, p. 155. Oxford: Basil Blackwell.
Dummett does not cite the author or title of the story, and this is probably because he did not remember them (see below for why I think this). So let me propose an identification. I am confident that it is ‘The Discovery of Morniel Mathaway’ by William Tenn, first published in Galaxy Science Fiction, October 1955. Tenn’s first-person narrator is indeed friends with an untalented artist who is proud of having “developed his own technique which he called smudge-on-smudge”. A purple box materializes in the artist’s studio and a time-travelling art critic emerges:
“Morniel Mathaway,” the man from the box said, “my name is Glescu. I bring you greetings from 2487 A.D. […] In my own period, I may say without much fear of contradiction, I am the greatest living authority on the life and works of Morniel Mathaway. My special field is you.”
According to the time-traveller, Mathaway is “one of the immortals the human race has produced […] the man with whom modern painting in its full glory, is said to have definitely begun”. But the critic is disappointed when he sees Mathaway’s actual work: to him it “looks like smears of paint on top of other smears of paint” and nothing like the reproductions he has brought with him in his book “The Complete Paintings of Morniel Mathaway, 1928–1996”.
Mr. Glescu materializes in Morniel Mathaway’s studio. Illustration by ‘Smith’.
Dummett seems to have forgotten some of the details of the story. First, he writes that the critic comes “from a century ahead”, but in the story he says that he is from “2487 A.D.” Second, Mathaway absconds in the time machine, and it is the art critic, stranded in the past, who assumes Mathaway’s identity and spends his life “producing the originals of the reproductions”. Third, the critic does not copy the paintings from the reproductions, but creates them himself, partly from memory but mostly from his expert sense of the Mathaway style which he had studied for so many years.
This last point, I think, is evidence that Tenn was writing for a readership of science-fiction fans who could be expected to be familiar with causal paradoxes in time-travel stories. Many of them must have read ‘By His Bootstraps’ by Robert Heinlein (writing as ‘Anson MacDonald’, Astounding Science Fiction, October 1941) and many other, less notable, works in the genre. This readership would be expecting an ending in which the critic was copying the paintings out of the book of reproductions, and so they could be surprised by the ending Tenn gave them, in which the critic “was actually painting the pictures”.
Anyway, these minor inaccuracies suggest that Dummett had summarized the story from memory, and could not recall the title or author. This is fair enough: in 1986 it was probably impractical for Dummett to track down a short story he had read in a science-fiction magazine thirty years previously. But let’s see what happened to Dummett’s summary of ‘The Discovery of Morniel Mathaway’ as philosophers took it up as a paradigm example of a causal loop. Surely someone would think to check?
In 2005, Michael Lockwood quoted Dummett’s summary of the story, and added this analysis:
Now if people ask ‘Where did the reproductions come from?’, you can give the perfectly good answer, ‘By being taken from the paintings?’ And if they then ask ‘Where did the paintings come from?’, you can answer ‘By being copied from the reproductions’. But there is no explanation for the cycle of events considered as a whole, no explanation for the conjunctive state of affairs, which comprises the existence of paintings and reproductions.
Michael Lockwood (2005). The Labyrinth of Time: Introducing the Universe, p. 174. Oxford University Press.
As we saw above, this is not correct, since the critic did not paint the originals by copying the reproductions. Had Lockwood been able to track down the story, he would have found the causal paradox to be subtler than the one he mistakes it for. Lockwood repeated this incorrect analysis in a 2010 book chapter with David Deutsch, again citing Dummett and not mentioning Tenn:
Another paradox, which often appears in science fiction, has been discussed by the Oxford philosopher Michael Dummett. An art critic from the future visits a 20th-century painter, who is regarded in the critic’s own century as a great artist. Thus, the reproductions exist because they are copied from the paintings, and the paintings exist because they are copied from the reproductions. Although this story threatens no contradiction, there is something very wrong with it. It purports to give us the paintings without anyone’s having to expend artistic effort in creating them—a kind of artistic “free lunch.”
David Deutsch & Michael Lockwood (2010). ‘The Quantum Physics of Time Travel’. In Susan Schneider, ed. Science Fiction and Philosophy: From Time Travel to Superintelligence. Wiley–Blackwell.
Probably 2005 was still too early to be able to easily find the story via corpus searches, but that can’t be said about the 2010s, when it became possible to find works like this using Google Books and other searchable online corpora. It took me only a few minutes to track it down: I did a Google Books search for
"time travel" "art critic" in 1986 or earlier (that is, using Dummett’s 1986 summary as the terminus ante quem), and the very first hit has the following snippet:
One of the neatest of these twist-in-time stories is “The Discovery of Morniel Mathaway” (1955), in which an art critic from the future comes to interview the twentieth century’s greatest artist.
Frank Northen Magill (1979). Survey of Science Fiction Literature: Five Hundred 2,000-word Essay Reviews of World-famous Science Fiction Novels with 2,500 Bibliographical References, volume 5, p. 2066. Salem Press.
Using the information in the snippet, it was straightforward to find the author and the text and confirm that the story matches Dummett’s summary. So it is disappointing to find philosophers in 2013 and later still relying on Dummett’s summary and not identifying the primary source. I’ll give two examples:
A clear lack of creativity, however, occurs in the story of a conceited, third-rate artist, whose tale leads to a knowledge puzzle. The story recounts an unfamiliar, strange event. A third-rate artist is visited by a time travelling critic from the future. In the art critic’s own time the artist is hailed as one of the greatest painters of the twenty-first century. When the artist shows the time traveller his paintings for his appreciation, the critic is disappointed by the poor quality of his work. He concludes that the painter cannot yet have entered his most creative period, in which he subsequently produces the celebrated masterpieces, for which he is famous in the critic’s own time. The art critic produces a portfolio of reproductions to show the painter what he means. Then the critic suddenly has to take leave to return to his own century. But the artist tricks him into leaving behind the portfolio of reproductions. The artist then spends the rest of his career producing the copies of the reproductions of the famous masterpieces. Thus he ensures his future celebrity.
Friedel Weinert (2013). The March of Time: Evolving Conceptions of Time in the Light of Scientific Discoveries, p. 244. Berlin: Springer.
Weinert cites the story to Dummett, Deutsch and Lockwood. Note that in Weinert’s summary the artist has become “third-rate” (not “fifth-rate” as in Dummett) and “one of the greatest painters of the twenty-first century” (not “twentieth”).
In a paper published in 1986, Michael Dummett envisages the following situation. A fifth-rate but conceited artist living the 20th century is visited by a time traveler from the 21st century, an art critic who is impressed by those of the artist’s works that have survived and have given him the reputation of being one of the greatest painters of his time. When the critic sees the works, he is disappointed by their quality, and realizes that the artist has yet to create the magnificent paintings that subsequently made him famous. The critic produces a portfolio of reproductions of the later works, but shortly after has to depart at short notice, being permitted for some reason to remain only for a limited time in the past. The artist manages to conceal the portfolio, and spends the rest of his life meticulously copying on canvas the reproductions left behind. These copies form the basis of his subsequent artistic renown.
Is there a problem here? Yes. […] The puzzle lies not in this, but in finding where artistic creativity enters the equation.
Storrs McCall (2014). The Consistency of Arithmetic and Other Essays, p. 217. Oxford University Press.
Note McCall’s phrasing, “Dummett envisages the following situation”, which removes the credit for the scenario from the anonymous science-fiction writer and implicitly assigns it to Dummett instead, thus further erasing the contribution of William Tenn.
I’ll give one final curious example.
There is a science-fiction story, set a couple of hundred years in the future, when time travel is assumed to be possible, about an art critic who becomes so fascinated by the works of a New York painter from our era that he travels back in time to meet him. The painter, however, turns out to be a worthless drunk who steals the time machine from him and escapes into the future; alone in the world of today, the art critic paints all the paintings that fascinated him in the future and had made him travel into the past. Surprisingly, none other than Henry James had already used the same plot: The Sense of the Past, an unfinished manuscript found among James’s papers and published posthumously in 1917, tells a similar story […]
Slavoj Žižek (2011). Living in the End Times, p. 30. London: Verso.
Žižek gives no citation for the story, but he can’t have got it from Dummett or any of the writers who derived from Dummett, because he doesn’t make Dummett’s mistake about who painted the originals. The detail that the painter is from New York is also correct (Tenn sets the story in Greenwich Village, a neighbourhood in Lower Manhattan). But Žižek is mistaken about the date (the art critic comes from five centuries in the future, not “a couple”), and about the painter being a “drunk” (there is no hint of this in Tenn). And the idea that The Sense of the Past has “the same plot” is laughable, unless you know so little about science fiction that you think all stories with time-travel are the same.
So what was Žižek’s source, if it wasn’t Dummett? Well, my best guess is that it was Žižek himself!
If this paradoxical structure is not clear, let us take another science-fiction example, William Tenn’s well-known story ‘The Discovery of Morniel Mathaway’. A distinguished art historian takes a journey in a time machine from the twenty-fifth century to our day to visit and study in vivo the immortal Morniel Mathaway, a painter not appreciated in our time but later discovered to have been the greatest painter of the era. When he encounters him, the art historian finds no trace of a genius, just an imposter, a megalomaniac, even a swindler who steals his time machine from him and escapes into the future, so that the poor art historian stays tied to our time. The only action open to him is to assume the identity of the escaped Mathaway and to paint under his name all his masterpieces that he remembers from the future—it is himself who is really the misrecognized genius he was looking for!
Slavoj Žižek (1989). The Sublime Object of Ideology, p. 57. London: Verso.
The details in this summary are all accurate. So in 1989, Žižek must have been able to read the story for himself, but by 2011 he had forgotten many of the details, and did not think to consult his own published work to check that he had got them right!
So what are the consequences, in this case, of relying on secondary sources, with no-one checking the primary source? First, Tenn doesn’t get the credit for the ideas in the story. The philosophical puzzle about where creativity comes from in a causal loop is right there in the original. Second, by never checking the original, philosophers have missed out on the sophistication of the ways causal paradoxes were treated in the science-fiction genre: by 1955, Tenn could take the idea of the causal paradox of creativity sufficiently for granted that he could subvert it in the last few paragraphs of his story. And third, the writers quoted above have perpetuated a mistake about the ending of the story. You might say that this mistake does not matter, since we can always imagine a hypothetical story that has the plot corresponding to the analysis. But if the identity of the story does not matter, then why use a real story as an example, why not make one up as needed? The implication, I think, is that a “found” philosophical puzzle is somehow more authentic or more valuable than one that is simply made up by the philosopher. But if that’s the case, then surely it’s important to get the details right.