Monday, March 17, 2014

Do scientists deliberately use technical expressions so they cannot be understood?

Secret handshake?
Science or gibberish?
“[E]xisting pseudorandom and introspective approaches use pervasive algorithms to create compact symmetries. The development of interrupts would greatly amplify Byzantine fault tolerance. We construct a novel method for the investigation of online algorithms.”

“[T]he effective diminution of the relevant degrees of freedom in the ultraviolet (on which morally speaking all approaches agree) is interpreted as universality in the statistical physics sense in the vicinity of an ultraviolet renormalization group fixed point. The resulting picture of microscopic geometry is fractal-like with a local dimensionality of two.”
IEEE and Springer recently withdrew 120 papers that turned out to be random generated nonsense and Schadenfreude spread among the critics of commercial academic publishing. The internet offers a wide variety of random text generators, including the one used to create the now withdrawn Springer papers, called SciGen. The difficult part of creating random academic text is the grammar, not the vocabulary. If you start with a grammatically correct sentence it is easy enough to fill in technical language.

Take as example the above sentence
“The difficult part of creating random text is the grammar, not the vocabulary.”
And just replace some nouns and adverbs:
“The difficult part of creating completely antisymmetric turbulence is the higher order correction, not the parametric resonance.”
Or maybe
“The difficult part of creating parametric turbulence is the completely antisymmetric resonance, not the higher order correction.”
Sounds very educated, yes? I have some practice with that ;o)The problem is that if you don’t know the technical terms you can’t tell if the relations implied by the grammar make sense. There is thus, not so surprisingly, a long history of cynics abusing this narrow target group of academic writing, and this cynicism spreads rapidly now that academic writing has become more widely available. With the open access movement there swells the background choir chanting that availability isn’t the same as accessibility. Nicholas Kristof recently complained about academic writing in an NYT op-ed:
“[A]cademics seeking tenure must encode their insights into turgid prose. As a double protection against public consumption, this gobbledygook is then sometimes hidden in obscure journals — or published by university presses whose reputations for soporifics keep readers at a distance.”
Kristof calls upon academics to better communicate with the public, which I certainly support. At the same time however he also claims professional language is unnecessary and deliberately exclusive:
“Ph.D. programs have fostered a culture that glorifies arcane unintelligibility while disdaining impact and audience. This culture of exclusivity is then transmitted to the next generation through the publish-or-perish tenure process.”
Let me take these two issues apart. First deliberately exclusive, and second unnecessary.

Steve Fuller, who is a professor for Social Epistemology at the University of Warwick, argues (for example in his book “Knowledge Management Foundations”) that the value of knowledge is related to the scarcity of access to it. For that reason, academics have an incentive to put hurdles in the way of those wanting to get into the ivory tower and make it more difficult than it has to be. It is a good argument, though it is hard to tell how much of this exclusivity is deliberate. At least when it comes to my colleagues in math and physics, the exclusivity seems more a matter of neglect than of intent. Inclusivity takes effort and most academics don’t make this effort.

This brings me to the argument that academic slang is unnecessary. Unfortunately, this is a very common belief. For example, in reaction to my recent post about the tug-of-war between accuracy and popularity in science journalism, several journalists remarked that surely I must have meant precision rather than accuracy, because good journalism can be accurate even though it avoids technical language.

But no, I did in fact mean accuracy. If you don’t use the technical language, you’re not accurate. The whole raison d’être [entirely unnecessary French expression meaning “reason for existence”] of professional terminology is that it is the most accurate description available. And PhD programs don’t “glorify unintelligible gibberish”, they prepare students to communicate accurately and efficiently with their colleagues.

For physicists the technical language is equations, the most important ones carry names. If you want to avoid naming the equation, you inevitably lose accuracy.

The second Friedmann equation, for example, does not just say the universe undergoes accelerated expansion with the present values of dark matter and dark energy, which is a typical “non-technical” description of this relation. The equation also tells you that you’re dealing with a differentiable, metric manifold of dimension 4 and Lorentzian signature and are within Einstein’s theory of general relativity. It tells you that you’ve made an assumption of homogeneity and isotropy. It tells you exactly how the acceleration relates to the matter content. And constraining the coupling constants for certain Lorentz-invariance violating operators of order 5 is not the same as testing “space-time graininess” or testing whether the universe is a computer simulation, to just name some examples.

These details are both irrelevant and unintelligible for the average reader of a pop sci article, I agree. But, I insist, without these details the explanation is not accurate, and not useful for the professional.

Technical terminology is an extremely compressed code that carries a large amount of information for those who have learned to decipher it. It is used in academia because without compression nobody could write, let alone read, a paper. You’d have to attach megabytes worth of textbooks, lectures and seminars.

In science, most terms are cleanly defined, others have various definitions and some I admit are just not well-defined. In the soft sciences, the situation is considerably worse. In many cases trying to pin down the exact meaning of an -ism or -ology opens a bottomless pit of various interpretations and who-said-whats that date back thousands of years. This is why my pet peeve is to discard soft science arguments as useless due to undefined terminology. However, one can’t really blame academics in these disciplines – they are doing the best they can building castles on sand. But regardless of whether their terminology is very efficient or not compared to the hard sciences, it too is used for the sake of compression.

So no, academic slang is not unnecessary. But yes, academic language is exclusive as a consequence of this. It is in that not different from other professions. Just listen to your dentist and her assistant discuss their tools and glues, or look at some car-fanatics forum, and you’ll find the same exclusivity there. The difference is gradual and lies in the amount of time you need to invest to be one of them, to learn their language.

Academic language is not purposefully designed to exclude others, but it arguably serves this purpose once in place. Pseudoscientists tend to underestimate just how obvious their lack of knowledge is. It often takes a scientist not more than a sentence to recognize an outsider as such. Are you be able to tell the opening sentences of this blogpost from gibberish? Can you tell the snarxiv from the arxiv?

Indeed, it is in reality not the PhD that marks the science-insider from the outsider. The PhD defense is much like losing your virginity, vastly overrated. It looms big in your future, but once in the past you note that nobody gives a shit. You mark your place in academia not by hanging a framed title on your office door, but by using the right words at the right place. Regardless of whether you do have a PhD, you’ll have to demonstrate the knowledge equivalent of a PhD to become an insider. And there’s no shortcuts to this.

For scientists this demarcation is of practical use because it saves them time. On the flipside, there is the occasional scientist who goes off the deep end and who then benefits from having learned the lingo to make nonsense sound sophisticated. However, compared to the prevalence of pseudoscience this is a rare problem.

Thus, while the exclusivity of academic language has beneficial side effects, technical expressions are not deliberately created for the purpose of excluding others. They emerge and get refined in the community as efficient communication channels. And efficient communication inside a discipline is simply not the same as efficient communication with other disciplines or with the public, a point that Kristof in his op-ed is entirely ignoring. Academics are hired and get paid for communicating with their colleagues, not with the public. That is the main reason academic writing is academic. There is probably no easy answer to just why it has come to be that academia doesn’t make much effort communicating with the public. Quite possibly Fuller has a point there in that scarcity of access protects the interests of the communities.

But leaving aside the question of where the problem originates, at prima facie [yeah, I don’t only know French, but also Latin] the reason most academics are bad at communicating with the public is simple: They don’t care. Academia presently very strongly selects for single-minded obsession with research. Communicating with the public, about one’s own research or to chime in with opinions on scientific policy, it is in the best case useless in the worst case harmful to do the job that pays their rent. Accessibility and popularity does for academics not convert into income, and even an NYT Op-Ed isn’t going to change anything about this. The academics you find in the public sphere are primarily those who stand to benefit from the limelight: Directors and presidents of something spreading word about their institution, authors marketing their books, and a few lucky souls who found a way to make money with their skills and gigs. You do not find the average academic making an effort to avoid academic prose because they have nothing to gain with that.

I’ve read many flowery words about how helpful science communication – writing for the public, public lectures, outreach events, and so on – can be to make oneself and one’s research known. Yes, can be, and anecdotally this has helped some people find good jobs. But this works out so rarely that on the average it is a bad investment of time. That academics are typically overworked and underpaid anyway doesn’t help. That’s not good, but that’s reality.

I certainly wish more academics would engage with the public and make that effort of converting academic slang to comprehensible English, but knowing how hard my colleagues work already, I can’t blame them for not doing so. So please stop complaining that academics do what they were hired to do and that they don’t work for free on what doesn’t feed their kids. If you want more science communication and less academic slang, put your money where your mouth is and pay those who make that effort.

The first of the examples at the top of this post is random nonsense generated with SciGen. The second example is from the introduction of the Living Review on Asymptotic Safety. Could you tell?


  1. In AWT the gradual convergence of theoretical strictly deterministic physics into vague and fuzzy postmodern philosophy (if not complete gibberish) is the undeniable consequence of Universe hyperdimensional geometry and its implicit emergent nature - and the physicists can do very little against it (they can still find a better job in time, indeed). It's like the description of scattering of ripples at the water surface at the sufficient distance from observer.

  2. Really, this form of criticism and the underlying assumptions should first target so-called legalese - it isn't quite the same thing when technical papers on a far out research limb use language that's perhaps less transparent than it could be to the outsider; it's a rather different thing when such language is meant to rule over the outsider's freedoms while putting in terms the outsider isn't meant to understand.

  3. Thank you for an excellent and very thoughtful piece.

    "Pseudoscientists tend to underestimate just how obvious their lack of knowledge is. It often takes a scientist not more than a sentence to recognize an outsider."

    This appears to have bypassed Zephir completely. As do most things.

  4. Yeah, on occasion I've wondered if 'Zephir' is a random postmodernism generator :p But at least for once his comment seems aptly in place.

  5. Kristof hasn't got the news that there's no royal road.

  6. "Could you tell?"

    Coming from a theoretical physical chemist grad student:

    For the first 'article', I don't know what many of the words mean, but I do know that "compact symmetries" has absolutely nothing to do with "Byzantine fault tolerance" so it is immediately suspect.

    For the second one, I DO know what "universality in the statistical physics sense in the vicinity of an ultraviolet renormalization group fixed point" means since I've studied the renormalization group in the context of the classical spin Ising model. So although I don't really understand the point they are trying to make, I can see how they flow from the first sentence to the third, so I can see how it would make sense if I had more background knowledge.

    Of course, since I'm not an expert in the field the probability of it being nonsense is still large, but not as large as the first paragraph.

    So it looks like to me that being an expert in one scientific field does give you some limited ability to sniff out garbage in other fields (especially if they are relatively close). However, it's very easy to overstep this and say something stupid, which is why most academics don't wander too far outside their fields if they care about their job.

    This is probably another reason most academics don't want to speak to the public too much. The public will inevitably ask questions outside the field the academic is comfortable talking about, and even if the academic is reasonably confident about their answer the probability of saying something stupid and hurting themselves is too high. Too much risk for too little reward.

  7. Even with cheating, professional jargon ("the phone book," Phys. Rev.) is impenetrable to outsiders. Insiders must know. The Sokal hoax and word salads are quality control. Production seeks quality assurance.

    European weather models are tested by forecasting backwards against known outcomes. American models are forward only. They fail (hurricane tracks). Economics, string theory, SUSY: "I had a guaranteed military sale with ED-209. Renovation program. Spare parts for 25 years. Who cares if it worked or not?" Robocop (1987).

    Professional management seeks pure numbers. Knowledge introduces social bias. The able must not be rewarded at the expense of the deserving. Everybody now has two sets of books, or they die. Knowledge is moribund, poisoned by very many very small grasping people.

  8. This comment has been removed by the author.

  9. Your just reliving the Sokal affair.

    If you don't want to linguistically tie language to mathematics and not share it with the public, of course, that is one's prerogative. But if you want research money, I would think you would want to be extremely clear about your proposals. To show some track record that the language you choose is fairly accurate to current research, as a writer, set the stage for your beliefs versus the framework and model you choose to advance perception regarding the areas you are working in.

    It's sort of like holding a standard(a motto perhaps) not just in the development of your science( some sign over your school doorway:) but to be as simplistically clear as possible, without denigrating that science. So are you saying it can't be done or that as a scientist are they are purposely being misleading?

    I think that would be to their detriment.

  10. Science's discovery, are like plots of land on some planet or moon?

    Just a side note to selling plots of land on mars there is an international treaty agreed upon about that:) as a question for some of your journalist fans. :)


    Plato's Nightlight Mining company:)

  11. Thanks MarkusM

    Bee you might want to chime in with your blog posts concerning this. You go way back too?:)

  12. "The PhD defense is much like losing your virginity, vastly overrated. It looms big in your future, but once in the past you note that nobody gives a shit."

    I'll add this to my collection of quotable quotes.

    There is a difference, though. In academia, rightly or wrongly, virgins aren't as highly valued.

  13. This is a good topic to ask what next after hype of such fundamental discoveries.

    Can anyone if we are immersed at a reduction to some idea of cause or history, some extention of a point in the E concept of energy transfer as explosion or doubled explosion that like a particular star our coherent view would see other persisting constellations from other stars?

    Can we make s program that would remove from a body of text that with scientific certainty would as logic remove (thus be said to falsify) papers that seem to be "pseudo-science "?
    Regardless of the nature of words.

    As replicative beings the struggle to defend territory as important as immigrants of the future or the masses filling historical space as a blindly lustful instinctually invasive species until there are no more frontiers or bases and castles not breached, or alpha male harums left in the care of eunuchs while corrupt empires fall or rise in turn.

    What is this CMB evidence but a bedsheet of blood as social proof of virginity at the consummation of the wedding day? Or the Higgs as forensic evidence in finer technology for who is the daddy?

    The word 'rhea' by the monotheistic God created the world in the ' beginning '. Biblical King James English. Or word did so as three dimensional quarks and dimensions 'logos ' , somewhat polytheistic New Testament interpretation.

    What, considering that so much is an unknown god or multiverse is still not understood (E &B its double forms and monopoles. ) that we creatures of space half or double values as a common conceptual mistake at the heart of half a deck errors in equations micro and macro that we still can only speculate what came before the Planck scale.

    Is cosmology or God an opponents in an advanced (to us) computer chess game? Or a pointless gamble where the bank always wins.?

    As we sacrifice our virgins to higher dimensional volcanoes and in the name of fertility say the unknown is non linear and hysteria, none of this swan song of Nobel Prizes represented by ladies, we blokes should ask ourselves if we are aware we are suffering from "old hymenizers " disease.

  14. Thank you again for an excellent article on a subject of importance. One further point: you don't need a specialist vocabulary to be turgid. Writing clearly is difficult, even when using everyday language.

    Another problem when communicating science to the unwashed is using everyday words which also have specialist meaning. What confusion has arisen by physicists using the word "theory"! Too late to fix now, but you guys should always be on your guard.

  15. Plato,

    No, I have no plans on covering the topic. I don't really see the point on repeating what has been reported on in other places already. I'm not a news outlet and have no intentions to become one. Best,


  16. Wild,

    Yes, that's a good point indeed. Sometimes the 'professional' terms also have an every-day meaning, which causes much communication problems. Theory isn't such a great example because even scientists don't exactly agree on what they mean with it, but take a word like 'significant' or the infamous 'data manipulation'. Best,


  17. Hi Bee,

    You know I hold you as much more then a news agency?:)I learn a lot from you, so I think your technical expressions came through quite nicely.

    Since E and B denote the relationship between the polarization amplitude and direction, warping due to lensing creates B-modes Secondary CMB Anisotropy PDF

    See: Wayne Hu Primer

    This should ring a bell? Perhaps, the ways in which one may look at the cosmos? This developing scenario is in a sense , an accomplishment not just in how far back we can go, but of how new tools brought to bear helped us to see the cosmos in interesting new ways.

    So the question might have been in relation too, "spectral geometry?"

    Quantum Gravity on a Quantum Computer?

    Shape from sound: toward new tools for quantum gravity

    Now most certainly, you must see why I would want a scientist like yourself to weigh in? :)


  18. How far back in time Bee?

    Surely, looking at the WMAP is more then looking at a 2d surface? Hearing the shape of the drum?

    If you want to look into the "internal structure of the sun" how might one want to do that?

    Thanks for everything Bee?

  19. Of and finally, I really liked your Listen to Spacetime....and hearing the shape of the drum.:)

  20. "No, I have no plans on covering the topic. I don't really see the point on repeating what has been reported on in other places already. I'm not a news outlet and have no intentions to become one."

    When I start my blog, what I certainly will not do is a post just mentioning something for which more detailed information is available elsewhere. I would only mention it if I had something to contribute. In general, it will be a high signal-to-noise blog. No pictures of cats. :-)

  21. We seem to parse languages and signals by various models similar to systems and interpretations of fundamental physics.

    Alphabet systems vary widely with few or many letters much like chromosomes dividing text or rules of syntax that bind the coherences of concepts as low space dimensions which sets then evolves space as a grounding first system of language.

    Before and beyond minimal units of meaning that seems to have periods of inflation yet questions of ultimate origins (minimal but turgid in compactifications or expansions adjacent to clearly coherent systems).

    An important idea is that an individual mind establishes language with words that work true to the inner logic so to understand as self systems before fundamental ideas can be communicated to others who may not be able to understand another's language and symbol system.

    Negative encounter of a systems evaluation is also useful to some degree. But without s careful stance in balance to wider model concepts we may see a text as turgid in its natural sense - a bloated dead animal on a hot summer day stinking and full of flies.

    By parallel analogy can we really say the expansion of space, galaxies as if dots on a balloon growing further apart
    unto isolation born and lost by light, I'd really the same turgid analogy as such isolation in a deeper acceleration of ideas as dark fluid cosmology and half sensible multiverses of inflation?

    But the spirit of confirmation of small steps in speculations foreseen is more productive for inquiry for it also affirms the value of all if us as human and the worth of individuals whose work is inquiry. I feel the joy and toast to this rare awakening joy of life.

    We can treat our conceptual languages in shorthands like numbered essential words but as we face a post alphanumeric and post economic awakening the one idea I was going to ask before this comment explosion of words was :

    What would be the minimum words (roots like time or space like words) that can convey and cover a shared but adapating language? I suspect these are surprisingly few.

  22. I agree that other than communicating a clear and original idea the vast referencing in a paper can be a waste of time and newsprint.

    Funding is all important (and praise to England in recent support for new collided experiments.) I simply do not think the cosmic observations make these projects pointlessly obsolete. But in offering a view that needs funding and access to one's own work some obscurification between how much to bid to peers between proposals from their researches released (even if the rationed release solely to renew funding) seems the current climate.

    The language problem far from originality can rapidly inflate what are reductions or expansions in systems that address symmetries.

    Basic English becomes long winded with few words and no verbs, quite idiomatic still needing proper names and technical terms. So some symmetry is lost in the translations.

    Or in a minimized irreducible subjective language system the symmetry can get long winded yet also short of formalized conceptual idioms lose something in the translation

    So instead of offering such a universally used sample of linguistic comparable text I will go directly to a casual translation from such a language hidden in the processes between translations. Can anyone recognize the terser passage translated from common English?

    (We as a part of a higher or unknown realm so assert transcending unities of Being emerged as faith or given eureka moments that there is an absolute in our living to which the best of life in our dreams of hope so obtain, where in this universe si ordains inflation's here thus hereafter; our brief interval sufficient for our needs, indifferent our shared frailty of errors that we need not in the main return to infinite darkness before our day.

    One consolation sums laws that life life goes on beyond us, our love and suffering unto our children.

    So in the song we sing fret not that the music may not return to sign. )

  23. I was really worried for a moment - I read the second one and thought it made perfect sense!

  24. You want to visit any number of different locations just once
    with the shortest distance to accomplish this.

    There. I translated the Google entry.

    No, readers insist no translation occurred. I agree. I was mistaken.


Comment moderation on this blog is turned on.
Submitted comments will only appear after manual approval, which can take up to 24 hours.