Elizabeth Holmes and the Language of Plausible Deniability

Daveena Tauber, Ph.D
7 min readJan 4, 2022

Staying out of jail was not among the list of reasons I typically gave my college students when trying to inculcate the value of writing clearly and persuasively. And yet the rhetoric-to-prison pipeline is very much on my mind as I contemplate the idea that Elizabeth Holmes fooled investors, journalists, and the public through rhetorical as well as technological sleight-of-hand.

It’s clear that Elizabeth Holmes left listeners at interviews, tech conferences, and investor pitches believing that her technology was running hundreds of tests on a single drop of blood, that those tests were run on Theranos machines, and that those machines were deployed on MedEvac helicopters (or hellavac as her partner-cum-conspirator once said). And it’s also true that listening to clips of her talks, it is surprisingly difficult to pin down the actual claims she made for her now infamous blood testing company Theranos. So difficult, in fact, that the slipperiness of her language has become part of her defense at trial.

While it appears that Holmes did sometimes lie point blank, she more often created verbal slippage between what her company was actually doing and what she said her company had made it possible to do, could do or had the ability to do. In a characteristic remark she says,

“We’ve made it possible to run comprehensive laboratory tests from a tiny sample or a few drops of blood that could be taken from a finger and we’ve made it possible to eliminate the tubes and tubes of blood that traditionally have to be drawn from an arm” [TED Talk 2014, emphasis mine]

As I binged the podcasts The Drop Out and Bad Blood: The Final Chapter, I found myself listening not only with the usual prurience but also with an ear for syntax honed by having spent an inordinate amount of time (let’s just call it ten thousand hours) reading and grading undergraduate papers. It’s an experience that tunes one’s ear not only to the many sentence-level issues that can interfere with clarity, but also to the rhetorical characteristics of — I believe this is the technical term — bullshit.

To be totally clear: giving and getting feedback on papers isn’t a picnic for anyone involved. To students it can feel like vindictive nitpicking — hence the need to make a larger case for the value of clear writing. For faculty, the cognitive friction of following student arguments through thickets of syntactic issues can make even the mildest-mannered occasionally fantasize about spilling oceans of red ink. Faculty take this work seriously not because they adore nitpicking, but because the ability to make and support clear claims lies at the beating heart of knowledge generation — the work Holmes claimed to be doing.

It is tempting, then, to read Elizabeth Holmes’s imprecise use of language as a symptom of her status as a drop out, of having had too few exhausted adjuncts plead “clarify,” “explain,” “show your evidence” across her work.

But the dreadful truth is that the rhetorical slippage of Holmes’s rhetoric was a feature not a bug. From the perspective of clarity, it was a linguistic failure, but it was one that enabled her success. If this sounds like a linguistic imprecision only a legal team could love, it’s very likely because her legal team did. Certainly, on the back end her legal team has pointed to these rhetorical moves as evidence that Holmes didn’t actually make the claims that her many interlocutors claimed they understood.

Softening or hedging claims is not always a nefarious move. In the humanities, where persuasive argument, not truth claim, is generally the goal, we teach students the seemingly paradoxical rule that a rhetorically weaker argument is often a conceptually stronger one. In other words, claims based on absolutes (all, every, never, always) are wide open to exceptions, whereas more limited or hedged claims (some, most, may, might, could) are harder to falsify. The problem is that Holmes applied hedging strategies not to making qualitative arguments, but to truth claims about her business and the science behind it. In being un-falsifiable, her claims were also unverifiable.

Holmes frequently helped create impressions in her auditors’ minds through eliding tenses. In an investor call played at trial, she mentioned early, exploratory work done with pharmaceutical companies and the military, but then conveniently neglected to mention that those contracts had long since ended. Likewise, she presented technological future states in ways that her audiences understood as referring to current functionality. At trial her lawyers argued unconvincingly that she had only ever been referring to the product vision.

Holmes also used phrases like “proprietary Theranos technology” deceptively. While her company may have been running blood tests on commercial blood analyzers in off-label ways she considered proprietary, there was no way for her listeners to know that she wasn’t referring to the Theranos Edison or Minilab machines she had demonstrated or described.

Holmes’s use of language failed both the central goals of your basic college paper: to make clear claims and to provide evidence. At trial, Shane Webber, Pfizer’s director of diagnostics, reported that during a phone call Holmes provided “oblique, deflective, or evasive non-informative answers to technical due diligence questions” and that “the nine conclusions in their summary document are not believable based on the information provided.”

Holmes was a college sophomore when she dropped out, and despite the sententiousness with which she presented her work, there was something notably sophomoric in her use of language. The repetition of oddly imprecise phrases like “tubes and tubes of blood” and the comparison of blood draws to “torture experiments” strike the ear of this erstwhile professor as slightly off in the way that student work so frequently is.

I am far from the first to notice something askew about Holmes’s discourse. In his New Yorker piece, Ken Auletta noted that Holmes’s descriptions of her own technology was “comically vague,” though he chalks this up to her being an engineer. While such cues were enough to rouse his suspicions, they were not quite enough to cut through the other signifiers of credibility that Holmes marshaled. But his article did catch the attention of John Carreyrou, the Wall Street Journal reporter who broke the story open.

In the English doctoral program where I trained to teach writing, there was a refrain: “language breaks down where thought breaks down.” In other words, language often becomes fuzzy when the underlying thought process becomes unclear. Another way to say this is that when the rhetoric doesn’t make sense, it’s worth pressing on the underlying ideas. Rhetorical literacy is not a magic decoder ring, but it is an important tool in life’s bullshit detector toolkit.

Unlike Carreyrou, who categorically suspected Holmes because of her age and lack of education (and sex?), I think there is actually a plausible version of the story where Holmes could have fit the mold of the life she envisioned for herself. She could have become a rigorous autodidact, accepted the guidance of experts, and practiced what Silicon Valley preaches about failure and iteration. Instead she and her COO attacked everyone who brought back a result they didn’t like, questioned their methods, or asked for clarification.

There’s no guarantee a college education would have prevented Holmes from committing fraud. Still, in one of my fantasy movie remakes, Holmes graduates from college, where she was required to take writing-intensive courses (even as an engineering major), goes on to found a company that advances medical technology, and uses her charisma to make clear and transparent claims about her technology rather than resorting to secrecy and obfuscation. In my fantasy she comes across as both entrepreneurial and erudite.

And since writing faculty don’t typically get to be the heroes of anyone’s story, in my other fantasy remake, a Stanford writing adjunct who is moonlighting at Theranos to support her teaching habit, jumps up, rips off her glasses (thus suddenly becoming beautiful) and demands, “state your claims, show us your evidence, clarify your tenses” — while Holmes’s high-powered board and investors look on, minds blown. IRL faculty do the next best thing: they deputize students to go forth and ask these questions themselves. I’m vicariously proud that it was a couple of recent college grads who saw that basic scientific principles were being violated at Theranos and called BS.

What troubles me about this story is not that Holmes herself dropped out (although I hope she’ll go back and complete her degree in or out of the big house); I’m troubled that her status as a drop-out is itself a status symbol in the Silicon Valley culture she dominated. The fact that her lack of formal education served as a humblebrag for her “natural entrepreneurialism” betrays the larger erosion of trust in education and in expertise. A fantastically gifted performer (or at least a fantastically coachable one), Holmes used her apparent gravitas and the transitive machismo of her board to present herself as a SME (subject matter expert) and a thought leader. In so doing she demonstrated the way these concepts themselves downgrade the figure of the expert from someone with agreed-upon training or expertise in a specific field or discipline to a person with an unspecified familiarity within a given “space.”

Holmes founded a unique lab for the testing of moral courage, and it is heartening to know how much of it her assays discovered in the many people who came forward to voice concerns or simply say, “I won’t be party to this.” I’d like to think that in a more rhetorically savvy world they wouldn’t have had to.

Looking back on my many years of teaching college writing, I find myself wishing that I had focused less on having students read and write about ethical issues and more on the question of what it means to use language ethically in our own lives.