Evidence Isn’t Everything or Everywhere

by Guest Blogger
October 5th, 2011

by Diana Senechal

I have nothing against evidence but am wary of the ascendant “evidence state.” I have seen nonfiction vigilantes marching around, asking, “where’s your evidence? where’s your evidence?” Those who question its rule get sneered out of town, if not steeply fined. This is not right. Evidence has its place, but it cannot and should not dominate everything. Even in the best arguments, evidence (strictly defined) is only one way of substantiating a point.

With the emphasis on “informational text” in the Common Core State Standards, students will be told, again and again, that they must back up their arguments with evidence. This is an important skill but only one of many. Some arguments use illustrative examples; some, reasoning; some, eloquence; and many, a mixture of all of these.

Arguments with evidence exist, of course, and can be quite convincing indeed. For instance, there’s Thomas Jefferson’s argument in the Declaration of Independence:

The history of the present King of Great Britain is a history of repeated injuries and usurpations, all having in direct object the establishment of an absolute Tyranny over these States. To prove this, let Facts be submitted to a candid world.

He has refused his Assent to Laws, the most wholesome and necessary for the public good.

He has forbidden his Governors to pass Laws of immediate and pressing importance, unless suspended in their operation till his Assent should be obtained; and when so suspended, he has utterly neglected to attend to them.

And so on. Jefferson offers fact after fact to demonstrate the “repeated injuries and usurpations” of King George III. In this case, evidence makes all the difference, since those considering revolt would want to be convinced that the King has indeed established absolute tyranny.

Some arguments use examples that illustrate rather than prove the point. In his letter “On the Shortness of Life,” Seneca (ca. 4 B.C.–65 A.D.) posits that our lives are long but that we make them short through “idle busyness”—that is, by occupying ourselves with empty activities. Here’s one example he provides:

Tell me, would you say that those men are at leisure who pass many hours at the barber’s while they are being stripped of whatever grew out the night before? while a solemn debate is held over each separate hair? while either disarranged locks are restored to their place or thinning ones drawn from this side and that toward the forehead? How angry they get if the barber has been a bit too careless, just as if he were shearing a real man! How they flare up if any of their mane is lopped off, if any of it lies out of order, if it does not all fall into its proper ringlets! Who of these would not rather have the state disordered than his hair? Who is not more concerned to have his head trim rather than safe? Who would not rather be well barbered than upright?

This is not “evidence”; rather, it shows vividly how the waste of time might play out. Anyone in his day might have complained that he exaggerated—that no one spent that much time at the barber’s or gave that much attention to their locks of hair. But that isn’t the point. This vivid hyperbole (if it is hyperbole) is there to convey the nature of the problem, not to prove its existence or extent. Beyond that, it’s just plain funny.

Then there’s argument with reasoning: for instance, in G. K. Chesterton’s essay “On Turnpikes and Mediævalism” (in All I Survey, 1933). Chesterton refutes a newspaper article’s assertion that a turnpike-gate with a toll is a relic of medievalism.

If we were really relics of mediævalism–that is, if we had really been taught to think–we should have put that question first, and discussed whether a thing is bad or good before discussing whether it is modern or mediæval. There is no space to discuss it here at length, but a very simple test in the matter may be made. The aim and effect of tolls is simply this:  that those who use the roads shall pay for the roads.  As it is, the poor people of a district, including those who never stir from their villages, and hardly from their firesides, pay to maintain roads which are ploughed up and torn to pieces by the cars and lorries of rich men and big businesses, coming from London and the distant cities. It is not self-evident that this is a more just arrangement than that by which wayfarers pay to keep up the way, even if that arrangement were a relic of mediævalism.

The logic is as follows: instead of fretting over whether a thing is modern or medieval, let us consider its merit. In this case, it seems fairer to have the roads maintained by those who ride than by those who cannot afford a car. Thus a turnpike-gate is more reasonable than the absence thereof, even if it is a relic of medievalism. Of course the reasoning is only part of the essay; the rest is wit and soul and a gift for putting nonsense in its place.

Eloquence, or compelling language, is its own kind of persuasion, usually but not always mixed with other kinds. Elizabeth Cady Stanton’s 1892 speech “Solitude of Self” stirs some kind of recognition and awe in the reader, through the very language.

And yet, there is a solitude, which each and every one of us has always carried with him, more inaccessible than the ice-cold mountains, more profound than the midnight sea; the solitude of self. Our inner being, which we call ourself, no eye nor touch of man or angel has ever pierced. It is more hidden than the caves of the gnome; the sacred adytum of the oracle; the hidden chamber of Eleusinian mystery, for to it only omniscience is permitted to enter.

The speech employs reasoning and example along the way; it argues that women need education equal to that of men, precisely because they must face so much of life alone. But the description of solitude, the core of her argument, has no evidence to prove it; it persuades through its starkness and beauty. If such argument were not permitted, simply because it lacked evidence, then we would lose a great deal of our nonfiction and fiction.

Each of these ways of substantiating argument has many variations, and they appear in combination more often than not. If schools are to give more attention to nonfiction (and to argument in particular), then they should acknowledge argument’s richness. If we assume that all good arguments must have evidence, then we narrow our view at great cost. This narrowness threatens even freedom of speech, as anything without evidence (or the appearance thereof) will be dismissed. We should resist such limitation. Argument thrives not only on statistics and facts, but also on uncertainties, questions, and risks; reasoning that startles you out of your assumptions; examples that make you laugh or shudder; and language that persists in the mind.

Diana Senechal’s book, Republic of Noise: The Loss of Solitude in Schools and Culture, will be published by Rowman & Littlefield Education in January 2012. She is the 2011 winner of the Hiett Prize in the Humanities, awarded by the Dallas Institute of Humanities and Culture.


  1. I love this piece, Diana, and the examples are wonderful. I’m a supporter of the CCSS for the simple reason that I think it passes the “Tiffany Test” — it makes it more likely, not less, that students like the ones I taught will get more meat in their diet and less thin gruel. But as Robert Frost did not say, something there is that does not love standards (his neighbor might have said good standards make good students).

    You piece is a good reminder–and we cannot be reminded too many times–that there are standards. And then there is the ability to soar above and beyond them.

    Comment by Robert Pondiscio — October 5, 2011 @ 8:25 am

  2. Well put, Diana. Reasoning and eloquence are not called for because that they can’t be measured by the pound and converted to a score by the readers at testing centers. If writing virtues can’t be quantified – in however dubious a fashion – they don’t count at the end of the grade.

    Comment by Howard Skillington — October 5, 2011 @ 12:22 pm

  3. I think you make a good point in reminding that there are many ways to show that something is important, many ways to make a sound argument. You also make a good point when you stress the fact that text (and by extent language) is not used or useful or interesting only when it is used to give evidence. But I do not believe that you make a convincing one for defending not using evidence to support our claims that can be supported by some. Funny enough, I believe we actually share that opinion.

    If someone was to implement a new program in schools, saying it will make students better at reading, you should ask for evidence. If he was to use “reasoning that startles you out of your assumptions; examples that make you laugh or shudder; and language that persists in the mind”, you could say “that is all good, but why do you believe that this program will make students better at reading?” You would be right; because that person is making a statement about the effect of something on students, an effect that can be verified. The only way to know if an empirical claim is true or not is to look for evidence. If we want real progress, we need to ask and look for it instead of just following what feels right, be it all beautifully presented. I doubt you would be against that.

    Saying that is not to say that literature, or philosophy (I teach it in a college) or art should be banned from schools. It is saying that we need to be honest in our arguments. If we say that teaching Milton is important because it makes students better at reading, we should look for evidence. If we cannot, or worse, if it was shown that it is counter-productive, we would have a very bad argument. But if we say that teaching Milton is important because it is a part of culture, or because it is beautiful, or because he writes about universal human themes, we would have a good argument. Because, lets repeat that to ourselves, education is not only making people able to read and to work; it is also widening their horizon and showing them things they won’t see elsewhere, things they will not seek by themselves.

    Saying that is neither to say that we should stop teaching everything that is not actually substantiated by hard evidence (much of finance and economy would take a solid blow…). Joking aside, there is a rule of thumb called “prudence”; there is also the hard reality that it is pretty hard sometimes to put an experiment to find evidence.

    Bottom line is : let’s not get all anti-evidence where it can (and should) be produced; this is how ideology rules instead of good practice. Instead, let’s question the assumption that only some goals are worthy in education. Let’s remind others that some goals are not the ones they are always asking and talking about.

    Jordan Raymond

    Comment by Jordan — October 5, 2011 @ 5:57 pm

  4. Jordan,

    Thank you for your thoughtful and challenging comment. I didn’t mean to imply (and don’t think I implied) that evidence should not be used where it’s needed. Clearly there are situations where evidence is in order.

    Now, where is evidence necessary, and where not? That question has been on my mind but takes a bit more space than I allowed myself for the blog (and I haven’t worked it all out yet).

    This much is evident, so far: assertions that can be proved with evidence (or at least backed up to a convincing degree) should have the evidence. If you say that “the poverty rate is rapidly rising,” you should provide figures to back up that statement. But if you say “poverty of the intellect is more dangerous than poverty of the wallet,” you must invoke something other than evidence, as the two kinds of poverty and danger cannot be measured against each other in any precise way.

    There’s much more to it–but as you yourself suggested, it comes down to what kind of question is being asked and what the assumptions behind it are. I welcome any more of your thoughts on this.

    Comment by Diana Senechal — October 5, 2011 @ 6:47 pm

  5. Excellent points Diana. The CCSS are all argument but no rhetoric, and it isn’t like there are no examples of standards embracing a fuller range of rhetorical approaches, they had to go out of the way to exclude them.

    The obvious reason is that computers can look for citations of evidence but not other kinds of persuasion.

    Comment by Tom Hoffman — October 5, 2011 @ 7:10 pm

  6. Diana,

    Of course, I did not think you would say that. But I think that saying “some things we do do not need to be supported by evidence” is going too far. We must accept that sometimes, the evidence we have is not good (sometimes even non-existent), that sometimes we do something to achieve a goal not usually brought up, tested or easily quantifiable (like widening students’ horizons). Even then, we should try to find some (Ben Goldacre make my point about that way better than I am : http://www.badscience.net/2011/05/we-should-so-blatantly-do-more-randomised-trials-on-policy/)

    About your example about “poverty of the intellect is more dangerous than poverty of the wallet”, I do not think that there can be no evidence or that you must invoke something else (well, you must always invoke something other than evidence when making an argument, at the very least reason and logic). On the contrary, we can search for evidence that can be of great help to settle the matter. First, you need to define what you mean by that. For example, if you meant something like “poverty of the mind makes people less likely to lead happy lives”, where you define happy lives by, say, the self-description people would give about the fulfillness, you can find evidence about the thruth (or falsity) of that claim by running some trials. If someone else say “poverty of the wallet makes people less likely to lead happy lives”, we can do the same. Then, comparing results, we would know better whatare the effects of both things on happiness.

    I know that my example is a bit shaky and that it does not stress important and very real difficulties about actually getting that evidence; my point is simply that even when you say something that seems to be impossible to support by evidence, usually because it appeals to values, you can (and should) define it in terms of observable effects; it then becomes clear that you can substantiate some evidence about it. “Being good” is basically too vague; explaining what you mean by that (make students like school more? Make them better at some task? Making them less likely to drop out? etc.) takes a lot of the vagueness out of the equation, making misunderstanding less likely and discussion a lot more productive.

    And by the way, thank you very much for you answer! I really like this blog. Congratulation for you prize and keep up the good work, you and everyone else on the team!

    Jordan Raymond

    Comment by Jordan — October 6, 2011 @ 1:20 am

  7. Excellent points: will they be on the test?

    Comment by Michael Fiorillo — October 6, 2011 @ 7:00 am

  8. Jordan,

    If we were to do what you propose, we would be required to turn everything into a science or social science. And I will fight that vigorously. Science is great when it does not replace poetry. And yes, there is poetry in nonfiction, and it should not be taken away.

    Plato’s writing is in many ways like poetry; for one thing, the arguments often depend on analogy. Here’s how many an argument of Socrates goes:

    We know that X is true about situation A.
    Situation A is analogous to situation B.
    Therefore, an analogy of X should be true for situation B.

    You can dispute any of this–but the hinge of the whole thing, the analogy, cannot be proved or disproved. You could invoke evidence to show that X is true about situation A. But it would clutter up the argument and wouldn’t make it more convincing in the end.

    As for your idea of conducting some trials to determine whether the poverty of the intellect detracts more from happiness than the poverty of the wallet, such a study would run into all sorts of problems. Are we fully aware of the extent of our happiness at a given time? Are our definitions of happiness stable, or is our understanding continually evolving? Do we readily disclose such matters to a stranger?

    But the uncertainty of these things should not prevent us from probing the nature of happiness, nor should it force us to invoke the social sciences when doing so.

    Comment by Diana Senechal — October 6, 2011 @ 9:38 am

  9. P.S. Thank you for the kind words–I didn’t mean to ignore them!

    (And I have no evidence for that except my own knowledge of my thoughts. Of course we can ask what intention is all about and what it means to “mean” to do something; that’s an interesting question in itself.)

    Comment by Diana Senechal — October 6, 2011 @ 10:06 am

  10. It’s in the nature of any argument that it cannot be reduced to, and it cannot ever be fully deduced from its evidence.

    To prove the second part – think of mathematical demonstrations. The teacher is at the whiteboard writing a proof for the theorem – any triangle ABC has sides |AB|=|AC| iff it has angles <(ABC)=<(ACB). (This is the first theorem one might think about isosceles triangles).

    What is the proof? In the first direction, pick the midpoint M of BC. The two ensuing triangles ABM and ACM have the corresponding sides equal. Therefore, by the edge-edge-edge triangle congruence criterion, the corresponding angles <(ABM) and <(ACM) are equal.

    For the opposite direction, pick M on BC so AM bisects the angle <(BAC). The two triangles ABM, ACM have corresponding angles and one side equal. Therefore, by the angle-edge-angle criterion, the corresponding sides AB and AC have equal length.

    What can be a more precise argument than this? But look! It is a sleigh of hand, because it is a recourse to the authority argument. I must know what I'm talking about, if I do not blink and talk fast. I have presented the 'data' about triangles to trick readers into believing me.

    But some sixth grader reading this may be familiar with isosceles triangles and congruence, and may say – no, there's no trick. The proof is correct!

    But it is still a sleigh of hand: for I have assumed the edge-edge-edge and angle-edge-angle criteria for triangle congruence: they have been presented in class the week before. They are to be taken as Axioms for our purpose.

    So are these believable axioms? To sixth graders, yes – again, by the authority of the teacher. To the more mathematically inclined – yes, because they are reducible to Euclid's axioms, and because it is assumed that the isosceles triangles theorem is within Euclidean theory.

    Case closed then? Have I reduced my argument to pure, unassailable data, with clearly delimited assumptions?

    All mathematicians would agree that what I've presented is a proof. This is – shall we say? – a 'cultural fact'. Nobody will speak against this proof. But some will observe that it is yet a sleigh of hand…

    And this brings us to the formal level. To make the proof completely unassailable, we need to write it formally down, with logical quantifiers, step by step changing each sentence and combining axioms and known theorems to get to the result.

    So at the meta-mathematical level, in Euclidean theory, the theorem can be proved. But, as I said that, I played the ultimate sleigh of hand – because this meta-mathematics is based on intuition. Or, if you will, it can be formalized with meta-meta-mathematics, which depends on meta-meta-meta-mathematics, all the way up to infinity.

    Because at a deep level any argument, however mathematical, will be based on intuition.

    And another thing to say here would be: why is the particular isosceles triangle theorem important? Do we know that because we are adepts of formal, based-in-fact arguments – or because of something else?

    Why did we pick this theorem, and not another? What is the point we're trying to prove with a little bit of 6th grade math?

    This is the point: no matter how 'mathematically correct' we are, it is still our taste, experience and fascination dictate which math we choose to do. That is why we've tried so hard to prove Fermat's last theorem and Poincare's conjecture about 3-spheres. That's why we treasure the mathematicians who proved them – Andrew Wiles and Grigori Perelman. It is a purely cultural phenomenon in the middle of the most matter of fact of the disciplines.

    Comment by andrei radulescu-banu — October 7, 2011 @ 9:00 pm

  11. Thank you, Andrei, for an excellent comment. (And it is my intuition, taste, and interest that leads me to call it excellent).

    I actually had a bout of insomnia afterward because I started working through theorems in my mind.

    I bring up the Poincare conjecture in my book, by the way. I also take a close look at theorem 1 in Newton’s Principia. I argue that we can get to more “big ideas” by looking closely at such a theorem (and some of the leaps in it) than by trying to work with “big ideas” themselves.

    But that’s another subject. Back to your comment: I appreciate your point that a mathematical proof involves a sleight of hand, even if it is explained all the way down to the axioms and known theorems (and even if each known theorem is explained).

    Comment by Diana Senechal — October 8, 2011 @ 7:33 am

  12. Thank you, Diana, for giving me new content I can use with my students, to help them discriminate between different tools for argumentation. The distinctions between evidence, illustrative example, reasoning, and eloquence will help them to put words on what they’re doing and what they can be doing. Our study of rhetoric through American history and literature will benefit from this categorization.

    And thanks for bringing up analogy as part of a critical thinking/analysis curriculum. I enjoy teaching the evaluation of analogy — in the American course as well as in my Philosophy elective — as part of argumentation. It’s something of my own devising, though, so I’m always gratified to be able to build on my amateur/autodidact experience with examples from thinkers such as yourselves.

    I’m not sure I agree with a fellow commenter’s dismissal of the CCSS, though. I’m pleased to see Common Core at least use the term “rhetoric” (not in any of my state’s pre-CCSS documents) and talk about argumentation without restricting the conversation to the use of “evidence”. And it talks — finally — about works of merit. Okay, it’s not perfect, but it’s a step in the right direction.

    Comment by Carl Rosin — October 8, 2011 @ 3:45 pm

  13. Thank you, Carl. I’m glad that you enjoyed the piece and the examples and am delighted that there’s something here that you can use with your students.

    And like you, I am not one to dismiss the CCSS, though I am wary of the current emphasis on “informational text” and how this may be interpreted.

    I wish there were a paragraph or two clarifying that literature comprises both fiction and nonfiction and that nonfiction at its best is in no way more serious, challenging, or important than fiction, poetry, and drama at their best. (Poetry is fiction and nonfiction at once, but that’s another matter.)

    In fact, I would recommend that the authors reconsider the very term “informational text,” which doesn’t seem quite right for philosophical works, speeches, sermons, and argumentative essays. Even historical and scientific works contain a great deal of analysis and commentary as well as information.

    As imperfect as the term “nonfiction” is, it allows for a wider range than “informational text”–especially if it’s understood that the line between fiction and nonfiction is fuzzy and that there’s much overlap. “Informational text” could then be a category within nonfiction.

    Comment by Diana Senechal — October 9, 2011 @ 11:58 am

  14. @Diana (#13): Beautifully stated, re non-fiction vs. informational. I’m going to copy this comment to pass it to administrators — your attention to the subtleties within the world of eduspeak is wanting in most Central Offices.

    Comment by Carl Rosin — October 9, 2011 @ 9:06 pm

  15. Diana,

    I am in no way trying to reduce everything to social sciences (or any other sciences). All I was saying was that if we want to know if it is true that a practice produces some effect, we should look for evidence proving that. And looking for evidence is the realm of sciences, not poetry. Again, this is not saying that poetry is not important, or that non-fiction should be bland dull instead of inspired.

    About the uncertainty of such matters as happiness (and a lot of others, even the ones that seems quite straightforward), I cannot agree more. I am not saying that we should abandon our probing. But we should use science to get evidence to shed light on them; so, in a way, I guess we should be “forced” to use it where need be. It cannot do everything, but all the questions you brought up can be answered this way : “Are we fully aware of the extent of our happiness at a given time? Are our definitions of happiness stable, or is our understanding continually evolving? Do we readily disclose such matters to a stranger?”

    Comment by Jordan — October 10, 2011 @ 11:55 pm

  16. Excellent post. I’ve noticed that students tend to approach evidence quantitatively: all evidence is good evidence, and more is better than less. What suffers is attention to argument, what we can do with evidence, or even the lack of it.

    Comment by Catherine — October 11, 2011 @ 11:06 am

  17. […] take care not to overdo it. More evidence does not automatically make for a better argument–nor do all arguments require “evidence,” strictly speaking. Machiavelli uses numerous historical examples to justify the points he makes in […]

    Pingback by Education Without “Stuff” | Diana Senechal — May 9, 2014 @ 4:03 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

While the Core Knowledge Foundation wants to hear from readers of this blog, it reserves the right to not post comments online and to edit them for content and appropriateness.