scientists took the customary route of commencing clinical trials.

When safety trials

suggested that the drug was relatively non-toxic, Pfizer ensued with second phase trails to
test for efficacy using volunteers afflicted with the heart condition. To Pfizer’s chagrin,
the trial results proved disappointing.

However, clinicians running the ongoing safety studies had learned of some interesting
side effects to the drug. As one of the original project team members recalled: “When the
first observations of erections were reported, the investigator who was doing the study in
Wales was a little bit embarrassed to mention it…When we explored the possibility of
treating patients with erection problems…some of those patients said ‘hey this is
working’, I need some more supplies…”2

Pfizer decided to commit another $340 million to the project by repeating Phase II
clinical trials using impotent men as a target group. The drug’s effectiveness was, this
time, beyond doubt. The Food and Drug Administration approved Viagra in March 1998.
It had generated revenues of US $300 million by early May 1998 in the US alone.
Ranked as the third best-selling drug of the past ten years, Viagra’s net present value is
estimated at over US$6 billion.

Clearly, Pfizer scientists were fortuitous in discovering its ‘accidental’ side effects in
toxicity trials. Yet why is this so surprising? After all, the role of luck in scientific and
technological advances is well recognized. At the same time, however, some scientists at
Pfizer had become increasingly aware of the likely efficacy of their drug when used to
treat erectile dysfunction, so much so that Pfizer was to loose its lucrative UK patent in
November 2000 for reasons of ‘obviousness’3.

Pseudo-serendipity by way of random variation: DNA

In other serendipitous discoveries, chance plays a much less significant role. For
example, elucidating the ‘double helix’ structure for Deoxyribose Nucleic Acid (DNA)
required a combination of unplanned events and a good deal of sagacity. James Watson
and Francis Crick were awarded the Nobel Prize for this discovery in 1962. Both were

13
frank about the spirit of adventure, tenacity, and youthful arrogance – but also frustration,
fear, mistakes, and serendipity – that marked this process of discovery. Having met at the
Cavendish Laboratory of Cambridge University in 1951, Watson and Crick recognized a
shared interest in the role of DNA in heredity. Neither had formally been assigned to
DNA research; in fact, both creatively masked some of their DNA work by relating it to
other in-house projects (e.g. the tobacco mosaic virus (TMV) and X-ray diffraction of
polypeptides and proteins). They intuited that DNA contained the ‘secret of life’, and that
its structure, “once found, would be simple as well as pretty” (Watson, 1999: 13). Aware
of the elaborate crystallographic analysis of Wilkins and Franklin at King’s College,
London, and the little progress it had achieved, Crick and Watson decided to approach
DNA elucidation instead through model building. In sharp contrast to Franklin’s
continued insistence that “there was not a shred of evidence that DNA was helical”, they
expected the molecule to contain two or three chains (Watson, 1999: 79). Watson’s
loosely related work on TMV seemed to support a helical structure. One night in June
1952, he was examining X-ray photographs of a new TMV sample and noticed its telltale
helical markings. If this were true of the virus, then why not also of DNA?

Further serendipitous events followed to direct Watson and Crick’s efforts. One entailed
a discussion over beer with John Griffith, a theoretical chemist, after an evening talk by
the astronomer Tommy Gold. Gold alluded to ‘the perfect cosmological principle’, to
which Griffith responded by wondering whether an argument could be made for a
‘perfect biological principle’, according to which genes would self-replicate. While this
hypothesis had been floating about for nearly three decades, most scientists assumed that
gene duplication required the creation of a ‘negative’ template from which a copy could
subsequently be crafted (Watson, 1999). Francis, by contrast, thought DNA replication to
involve attractive forces between the flat surfaces of the bases with different structures.
That evening they agreed that Griffith would try to calculate these attractive forces.
Shortly afterwards, Griffith responded by hinting that adenine and thymine should stick
to each other as, in principle, should guanine and cytosine (the four bases of DNA).
Francis paired this observation with prior research by Chargaff, suggesting that these four

14
bases seemed to occur in equal quantities. This suddenly made sense. As Crick
commented:
The key discovery was Jim’s determination of the exact nature of the two base
pairs (A with T, G with C). He did this not by logic, but by serendipity…In a
sense Jim’s discovery was luck, but then most discoveries have an element of luck
in them. The more important point is that Jim was looking for something
significant and immediately recognized the significance of the correct pairs when
he hit upon them by chance (Crick, 1988: 65-66; italics in original).

Some time later, yet another non-trivial idea surfaced. “It came while I was drawing
fused rings of adenine on paper. Suddenly I realized the potentially profound implications
of a DNA structure in which the adenine residue formed hydrogen bonds similar to those
found in crystals of pure adenine” (Watson, 1999: 145). Thus, Watson concluded, DNA
might consist of two chains with identical base sequences held together by hydrogen
bonds between pairs of identical bases Even if this hypothesis subsequently proved to be
false, it did point Watson towards the correct ‘double helix’ structure for DNA.

Perhaps the most significant serendipitous event was the sharing of an office with Jerry
Donohue, a crystallographer who, when confronted with Crick and Watson’s semi-
developed model, claimed that most textbooks representations of tautomeric forms of
guanine and thymine were highly improbable. This proved to be a terrible
disappointment, as they relied on these forms in building their model. As Watson
recollects: “Thoroughly worried, I went back to my desk hoping that some gimmick
might emerge to salvage the like-with-like idea. But it was obvious that the new
assignments were its death blow” (Watson, 1999: 151). Only the next morning, when
clearing his desk of papers, did Watson conceive of a sequence of adenine-thymine pairs,
held together by two hydrogen bonds, as being identical in shape to a guanine-cytosine
pair, held together by at least two hydrogen bonds. When verifying this possibility,
Donohue raised no objections. Thus, the double-helix structure was conceived.
The unforeseen dividend of having Jerry [Donohue] share an office with Francis,
Peter and me, though obvious to all, was not spoken about. If he had not been

15
 with us in Cambridge, I might still have been pumping for a like-with-like
structure. Maurice [Wilkins], in a lab devoid of structural chemists, did not have
anyone about to tell him that all the textbook pictures were wrong (Watson, 1999:
163).

The first in a series of four articles announcing the discovery was published Nature in
1953. The only allusion to its considerable implications for heredity was famously
entailed in its final sentence: “It has not escaped our notice that the specific pairing we
have postulated immediately suggests a possible copying mechanism for the genetic
material” (Watson and Crick, 1953: 737). The article was just over a page in length.

Pseudo-serendipity as the unintended consequence of design: PCR

Kary Mullis was awarded the 1993 Nobel Prize in chemistry for his discovery of a DNA
amplification method called polymerase chain reaction (PCR), a molecular biological
method for creating multiple copies of a specific stretch of DNA (e.g. a single gene or
part of a gene) without using a living organism. By taking DNA polymerase from heat-
loving (thermophylic) bacteria, Mullis was able to separate DNA in two strands by
heating it to 96°C, but without also destroying the DNA polymerase used to duplicate
DNA when cells divide. Thus it was no longer necessary to manually add DNA
polymerase after each DNA strand separation. Instead, the process could be simplified
and automated to copy as many DNA strands as required. This ingenious method is used
today in diagnosing infectious and hereditary diseases, and in DNA fingerprinting.

Mullis came to this discovery one night while driving his car along California’s Highway
128. At the time (in 1983), he was employed at the Cetus Corporation, who was to sell
the rights to Mullis’ PCR method to Hoffmann-La Roche for US$300 million. With time
on his hands, he had started to think about an improvement in DNA sequences, a thought-
puzzle that ultimately lead him to PCR. Mullis recognized the importance of his
discovery immediately: “This simple technique would make as many copies as I wanted
of any DNA sequence I chose, and everybody on Earth who cared about DNA would

16
want to use it. It would spread into every biology lab in the world. I would be famous. I
would get the Nobel Prize” (Mullis, 2000: 6-7).

Despite the significance of Mullis’ discovery, it was not entirely unexpected given that
PCR relies on a reconfiguration of well-known, existing technologies. As Fields (2001:
10052) points out, “PCR arose from dideoxy sequencing, developed in Frederick
Sanger’s laboratory about 6 years earlier”. In turn, dideoxy sequencing was based on the
‘plus and minus’ system, another of Sanger’s techniques. The ‘plus and minus’ system
itself was enabled by various pre-existing technologies, including radioactive precursors
to follow DNA molecules, other methods for separating DNA fragments, the isolation
and characterization of DNA polymerase, and various other techniques (Fields, 2000).
So by the early 1980s, all of reagents and procedures were in place for PCR to
come about. Many molecular biologists other than Kary Mullis could have
invented PCR, making its eventual introduction inevitable. All that was needed
was the inspiration of one individual with the willingness to putter about with
enzymes and primers (Fields, 2001: 10052; italics added).

Indeed, in his autobiographical account Mullis acknowledged that “there was not a single
unknown in the scheme. Every step involved had been done already” (Mullis, 2000: 9).
Mullis’ sagacity resided not in seeing what no one had seen before, but in thinking what
no one had yet thought of.

Serendipity revisited
Despite individual differences, these four innovations share at least one important feature:
they were made by individuals able to ‘see bridges where others saw holes’. To see
bridges, or ‘matching pairs’, is to creatively recombine events based on the appearance of
a meaningful rather than causal link. Both Fleming and Pfizer’s scientists applied
creativity and practical judgment in matching observations of unforeseen events with
findings reported by others, and in selecting which of these combinations might be
fruitful. They rightly interpreted coincidences as meaningful in the context of the
knowledge available to them at the time. However, the particle from the mycology labs

17
wafting through Alexander’s Fleming’s open window to contaminate a bacterial culture
is a random variation, as were the unusual changes in temperature. By contrast, the
unanticipated side effects of sildenafil citrate surfaced in part as a result of research
design; after all, toxicity trials tend to use men between the ages of 18 and 30, as did
Pfizer’s clinical trials.

Crick and Watson’s discovery of the ‘double helix’ structure of DNA was marked by
various unplanned events such as Watson’s loosely related work on TMV (corroborating
their suspicions of a helical structure), and exchanges with Griffith and Donohue (in
directing them towards the specific, but unorthodox, pairing of bases). Yet they always
knew they were after the structure of DNA, believing it to contain the secret of life. Thus,
DNA illustrates pseudo-serendipity, insofar as chance events enabled the unraveling of
the molecule, yet these events never caused them to deviate from this original target.
Similarly, Mullis had been searching for a method that would improve the replication of
DNA fragments. His PCR method relied entirely on existing technologies. Eccentricity,
rather than chance, may have played a role in Mullis’ discovery. A product of 1960s
Berkeley, he confessed to far-reaching experiments involving LSD, antihistamines, and
various home-made psychoactive drugs, using himself as the principal subject. Based on
his autobiographical account, Mullis’ innovation rather seems to have been a
consequence of sloppy research and naivety (“in truth, I was terribly naïve…if I had had
more knowledge about what I was doing, PCR would never have been invented”, p. 24),
a refusal to be bound by prevailing scientific paradigms, and a keen eye for matching
pairs of technological developments, able to select what seemed like effective
technological combinations; a feat accomplished by leveraging his powers of
observation, combination, adjudication, and application.

While the ‘unplanned’ or ‘random’ element retains a role in serendipitous innovation, it

appears to be more peripheral than typically assumed. For it is but one of several raw
materials – albeit an important one – on which serendipity relies but is not, strictly
speaking, tantamount to. Chance is an event, serendipity a capability. Chance has no bite
but for creativity in brokering random events meaningfully with other events.

18
“Significant inventions are not mere accidents”, suggested Nobel laureate Paul Flory
(Roberts, 1989: x). “Many a man floated in water before Archimedes; apples fell from
trees as long ago as the Garden of Eden […] chance discovery involves both the
phenomenon to be observed and the appropriate, intelligent observer” (Walter Cannon, as
quoted in Merton & Barber, 2004: 171-2). As Jung (1973) points out, the process of
recognizing a-causal events is subjective. It is intrinsic to the observer. As aptly put by
Willis Whitney, formerly a director of research for General Electric:
In every individual’s stock of knowledge (his conscious and subconscious assets)
there lie the peculiar items or records of his former thoughts. Some of them may
‘pop out’ or ‘come to mind’ when a novel or unexpected event crosses his mental
threshold. Some sort of catalysis has taken place. This all indicates dependence of
the gift of serendipity upon the total (even forgotten) knowledge and training of
the individual (quoted in Merton and Barber, 2004: 173).

A typology of serendipity
By exploiting the similarities and differences across these case histories, we are able to
arrive at a typology of serendipity. When it came to the discovery of PCR and DNA,
those involved found what it was they were looking for but by way of chance. By
contrast, in the discoveries of sildenafil citrate and penicillin, scientists discovered
something different from what they were looking for. The former can be labeled ‘pseudo-
serendipity’ (Roberts, 1989), also known as ‘serendipity analogues’ (de Chumaceiro &
Yaber, 1995). Here the objective remained unchanged, but the route towards achieving
this objective proved unusual and surprising. By contrast, the latter is ‘true serendipity’,
or ‘serendipity proper’ (de Chumaceiro & Yaber, 1995), so as to emphasize a change in
objective as a result of the discovery process.

A further distinction can be made between chance as the unintended consequence of

research design, and chance as pure random variation. So, for instance, in the sildenafil
citrate and PCR examples, opportunities arose as a direct consequence of the way the
study was designed: the unintended side-effects of sildenafil citrate appeared precisely
because Phase 1 clinical trials generally use healthy male volunteers (rather than

19