Maxim: Figures don't lie, but liars do figure (1884 February 29)

Sat Apr 10 11:59:16 UTC 2010

Several comments. The message came out awfully long and I had to break
it down into three parts. I am also only including a part of Garson's
original post.

Part I.

First, a thought on the 1889 citation from a C. D. Wright speech.*

Just as Garson did, I stumbled across the ASA volume.

http://bit.ly/9gg7e5
(American Statistical Association, 1889)
Convention of Commissioners of Bureaus of Statistics and Labor.
Hartford, June 25, 1889. p. 283 [43]
> It has been said that figures will not lie. It is equally true that
> liars will figure. It is our duty to prevent liars from figuring in
> the interest of any theory, by presenting original data fairly.

However, I interpret the text differently. Judging from the first and
last grafs immediately bracketing the "speech", I don't believe that
this is an "excerpt" from the address, as Garson has concluded. My
interpretation is that this is a telegraphic report on the content of
the speech--notes, if you will. So the text is not faithful to the
original. In this context, I would trust the second text that Garson
cited as being a more accurate representation of the actual address.

Our searches, unsurprisingly, overlapped again. At first, I did not
notice that Garson used The Sources of Our Population piece in The New
Englander (Aug. 1852), so I was going to refer to it anyway.

http://books.google.com/books?id=cWdJAAAAMAAJ&pg=PA393
The New Englander. Vol. 39. Aug. 1852
Art. IV.--The Sources of Our Population. p. 393
> The old saying, that " figures won't lie," is true, without doubt; and
> the same may be said of letters, marks, and other signs of thought.
> But the mode in which many use figures, in order to carry a point, has
> sometimes tempted us to believe that the hasty remark of the Psalmist,
> if paraphrased thus—" all men"—who deal in statistics—" are liars,"—is
> not far from the truth.

I like the sound of "all men who deal in statistics are liars".
Unfortunately, the source cites no authors or editors (save for a couple
of articles that are reproductions of speeches).

But I do have something to add. The same entire piece had been reprinted
in The Foreign Evangelical Review; Vol. 1:3; Edinburgh [London]:
November 1852; p. 650. [The entire volume is listed under The British
and Foreign Evangelical Review and Quarterly Record of Christian
Literature.] This means that the piece appeared on both sides of the
Atlantic.
http://books.google.com/books?id=pfkDAAAAQAAJ&pg=PA650

Another related line, comes from the mouth of a preacher at Boston's
West Church.

http://books.google.com/books?id=xFENAAAAYAAJ&pg=RA15-PA47
In Remembrance. Address, on the Occasion of the Death of Charles Greely
Loring, Delivered in the West Church, October 20. Boston: 1867
Sincerity. p. 47-8
> /Figures/ cannot lie, we say. In the wrong rows they are the greatest
> liars in the world, as every accountant and swindled trader learn.
> Facts are /stubborn/ things, and very misleading.

[GB bundled several West Church pamphlets in the same record--the
listing is under The Hand of God in the Great Man: A Sermon Delivered in
the West Church Boston, Occasioned by the Death of Daniel Webster. By
Cyrus Augustus Bartol. Boston: 1852]

The same sermon had been reprinted as "An Address before the Essex
Conference, Feb. 28, 1872" in The Radical, Vol. 10, May 1872; pp. 351-2.
http://books.google.com/books?id=SgqgAAAAMAAJ&pg=RA1-PA351

I was initially amused by the coincidence of dates, but this was quickly
dispelled by the realization that the GB volume contained many pieces.

The Massachusetts Teacher volume from 1869 contains two articles that
may also serve as marginal precedents.

The first is devoted to the question of how figures can "lie" in its
entirety.

http://books.google.com/books?id=2dQBAAAAYAAJ&pg=PA138
The Massachusetts Teacher. Volume 22 (3d series, vol. 4):4. April 1869
Editors' Department. Figures Do Sometimes Lie. p. 138ff
> "Figures don't lie!" So everybody says, and what everybody says ought
> to be true. And yet figures do sometimes tell wrong stories. Not that
> they are to blame. ... It is not always wise to trust to the figures
> as recorded on paper ; for it is just as easy for the pen to make 5 as
> 3, and carelessness or dishonesty will sometimes dictate that one be
> made for another.

The story goes on in the same vein for 5 pages.

The Massachusetts Teacher. Volume 22 (3d series, vol. 4):7. July 1869
A word about Statistics. p. 244
> Again, question should be framed so as to be free from ambiguity, and
> so as to admit of accurate answers. The neglect of this principle has
> been prolific in inquiries almost sure to draw forth very different
> results from kindred facts. A remedy is sometimes sought by giving
> directions evidently arbitrary and calculated to /insure/ inaccuracy.
> In this way, at great expense of labor, an immense amount of falsehood
> is put into figures, all the more deceptive because of the somewhat
> prevalent conviction that "figures will not lie."

Now, the former does have a play on "figures don't lie", but it is
rather internal--there is no reversal that makes the phrase in question
so entertaining. The latter does not even suffer that far--it simply
points to the falsehood based on the statement that "figures don't lie".

There is a different kind of reversal in the 1886 Popular Science
Monthly. This already postdates Garson's find, but it's still worth
mentioning because the contention here is, in reference to statistics
that "figures always lie".

http://books.google.com/books?id=hdrUAAAAMAAJ&pg=PA794
> But, in spite of all defects, two special merits belong to Le Play's
> work in social science. The first is, that he insisted on studying
> concrete, not abstract "society" ; he employed the statistical method.
> It has been said, to be sure, that "figures always lie" ; and
> certainly charts and diagrams, and brace-synopses that profess to set
> forth social facts, either past or present, should be accepted with
> profoundest caution. They are things to be used as Spencer used them
> in his "Descriptive Sociology." not as being themselves final results,
> but only as a means that may help us in reaching results more nearly
> final.

1886 seems to be a banner year for doubt.

http://books.google.com/books?id=g_wKAAAAIAAJ&pg=PA443
Papers of the American Historical Association. Volume 1:6. 1886
Meetings of the American Historical Association. Secretary's Report.
Evening Session. September 8, 1885. Professor Emerson's Paper. p. 23 [443]
> All this sounds very attractive. The average man accepts without much
> question the assertion that facts will not lie. And yet every one who
> has to deal with facts knows that nothing, unless it be figures, will
> lie so completely as facts. The question in every case of historical
> narration is not, "Has the man got facts?" but "What facts has he got,
> and how does he put them together ?" There is never a controverted
> point on any subject upon which a vast array of opposing facts cannot
> be summoned to the argument.

A similar doubt in figures is expressed in an 1880 piece as well:

http://books.google.com/books?id=3LU0AAAAIAAJ&pg=PA105
The Medical Jurisprudence of Insanity. By J. H. Balfour Browne. 2nd ed.
San Francisco: 1880 [copyright 1875]
Chapter II. On the Causes of Insanity. pp. 104-5
> We are glad to see that Dr. Ray, in an able article upon the causes of
> insanity, has expressed his dissent from the popular adage that
> figures wont [sic] lie, and shows that it is upon far other grounds
> that we must expect to arrive at correct conclusions as to the causes
> of mental disease.

Is this a bizarre reversal to go from "figures don't lie" to "figures
always lie"? Not so much. What is it that has prompted the reversal from
the classical "figures don't lie" to not only the claims "figures always
lie" but also to qualifying puns "Figures don't lie, but liars
[sometimes] figure"? Wiki article on the History of Statistics barely
mentions Quetelet (1796-1874) and omits William Playfair (1759-1823)
entirely. Quetelet introduced "social physics" and the concept of "the
average man" (the one who "has one breast and one testicle", as someone
later quipped). Playfair is responsible for the introduction of charts
("figures") into sociology and economics. Where "figures" originally
referred only to numbers, in the 19th century it referred to statistical
charts as well after Playfair. Le Play (1906-1882), mentioned above,
also made a major sociological contribution by publishing a series of
analyses of family budgets. To that one must add the rise of Darwinism
and, in particular, positivism--as Comte's Course, which resurrected the
near-scientific concept of sociology, was first published in English in
1865. Comte, in particular, predicted emergence of sociology as a hard
science (a prediction not yet realized). The influences of Darwinism and
Positivism on language (and on much else) are far too frequently
underestimated. It is the interplay of all these elements that would
have contributed to the doubt in numerical data that was a prerequisite
for the coinage of the pun on figures and liars. The 1880s are the
infancy stage of mathematical statistics, but it is also period of
maturation of data-driven sociology and economics. Early 19th century
statisticians and economists would have put near absolute faith in
numerical data--the kind of faith that generated a backlash exemplified
by the sources above.

It would still take the wit of someone in the middle of this debacle to
come up with a flying defense of numerical data and to blame the critics
(the "liars") for the misrepresentations of data (as exemplified by the
last three citations). In this climate, it makes perfect sense (to me,
at least) that the phrase "Figures don't lie, but liars do figure" would
have appeared in the 1870s or early 1880s. This is one reason why some
late 20th century quotation collectors try to pin the phrase on Mark
Twain. (I kid you not!) The best part is that the joke can be used both
in support of data-driven methodologies (which is how Wright used it)
and in opposition to them (see below for the anti-Wright quote).

Of course, my claim would be proven wrong if there is an instance of the
same phrase (not an antecedant) appearing prior to 1865. On the other
hand, if there is a source dealing with sociology or economics that uses
the same phrase and predates the 1884 cite, it would go a long way to
confirm my theory.

What is missing in the 1880s citations that Garson found
(disclaimer--not his fault, it is inherent to the source) is the
underlying increased reliance on numerical data in driving industry and
commerce. In fact, in this whole collection, it is only the 1888 LATimes
article that stands out as incongruous (and more in line with the 1869
Massachusetts Teacher articles, cited above). But once the phrase was
out of the bag, it certainly could not have been contained.

This is about all I have to add on precedence.

        VS-)

PS: *A slight correction of Garson's description of Carroll D. Wright.
He was the President of ASA and the very first Commissioner of Labor at
the time when he made the 1889 speech. He was placed in charge of the
Census in 1893 and became the President of AAAS in 1903, before stepping
down as Commissioner of Labor in 1905. His near 20-year tenure in that
post may serve to explain the 1904 barb directed at him.

On 4/9/2010 9:25 PM, Garson O'Toole wrote:
> Figures don't lie, but liars do figure.
>
> This remark is sometimes attributed to Carroll D. Wright who was a
> prominent statistician, former U.S. Commissioner of Labor, and former
> President of the American Association for the Advancement of Science.
> Wright did use a version of the phrase in 1889 at a convention;
> however, the proverb appeared earlier in a newspaper in 1884 without
> attribution. In addition, Wright called the second half of the proverb
> a new saying when he used it, but did not suggest that he coined it.
> Specifically in 1889 he said, 'The old saying is that "figures will
> not lie," but a new saying is "liars will figure."'
>
> ... I was strict in requiring the four
> terms that capture the wordplay: figures, lie, liars, and figure.
>
> The oldest three citations I could locate occur in the years 1884 and
> 1888. No attributions are given in these early cites. In the first
> instance the quotation follows other aphoristic remarks and prefaces a
> story about successful and unsuccessful merchants.
>
> Citation: 1884 February 29, Grand Forks Daily Herald, The Fire
> Laddies: The Great Public Benefit, Page 1, Column 2, Grand Forks,
> North Dakota. (GenealogyBank)
>
> "Keep pushing, 'tis wiser than sitting aside, And sighing, and
> waiting, and watching the tide; in life's earnest battle they only
> prevail, Who daily march onward and never say fail."
>
> Figures don't lie, but liars do figure. The continual complaining of
> some merchants make everybody tired. If the chronic kickers will pay
> more attention to their own business they might in time accomplish
> something for themselves. but no, they complain about dull trade and
> wonder at the grand success of the great sale of bankrupt stock now
> going on at the St. Paul Store.
>
>
> In the second citation in 1888 the quotation is spoken by Col. L. F.
> Copeland during a speech attacking the ideas of the famous freethinker
> Robert Ingersoll.
>
> Citation: 1888 July 13, Los Angeles Times, Long Beach: Summer Training
> School and Alliance Assembly: The Mistakes of Ingersoll Pointed Out by
> Col Copeland, Page 2. (ProQuest Historical Newspapers)
>
> Mathematics is an exact fact; figures don't lie, but liars sometimes figure.
>
> http://news.google.com/archivesearch?q=%22sometimes+figure+Bob%22&
>
>
> In the third citation the maxim is used in an argument about free trade.
>
> Citation: 1888 October 2, Sacramento Daily Record-Union, Page 2,
> Column 3, Sacramento, California. (Chronicling America: Library of
> Congress)
>
> It was a highly protective measure. The cry of free trade was a false
> one, and was maliciously put forth by "the uncrowned king" and other
> Republican leaders. Figures would not lie, but liars will figure, and
> were doing so in this campaign.
>
> http://chroniclingamerica.loc.gov/lccn/sn82014381/1888-10-02/ed-1/seq-2/;words=lie+Figures+liars
>
>
> On June 25th of 1889 the statistician Carroll D. Wright gave the
> opening remarks at a Convention of Commissioners of Bureaus of
> Statistics of Labor. Two different versions of his remarks were
> published, and I present excerpts from both below. Wright uses the
> proverb in both versions. I think that the wording used in the
> citation immediately below indicate that Wright was not the creator of
> the aphorism. He says it is a "new saying", but this suggests that it
> was already being disseminated.
>
> Citation: [1889 June 25, date of opening remarks at  the convention]
> 1890, Joint Documents of the State of Michigan for the Year 1889:
> Volume III Part I., Seventh Annual Report of the Bureau of Labor and
> Industrial Statistics Issued February 1 1890, Page 311, Darius D.
> Thorp: State Printer and Binder, Lansing, Michigan.  (Google Books
> full view)
>
> Excerpt from the opening remarks of Carroll D. Wright:
> The old saying is that "figures will not lie," but a new saying is
> "liars will figure." It is our duty, as practical statisticians, to
> prevent the liar from figuring; in other words, to prevent him from
> perverting the truth, in the interest of some theory he wishes to
> establish. We can only do this by being absolutely fair ourselves. But
> the limitations of which I speak almost prevent fairness and justice
> on the part of the statistician.
>
> http://books.google.com/books?id=20pOAAAAMAAJ&q=liars#v=snippet&
>
>
> Citation: June 1889, American Statistical Association, Convention of
> Commissioners of Bureaus of Statistics of Labor, Page 283, Vol. 1, No.
> 6, Publication of the American Statistical Association.  (Google Books
> full view)
>
> Excerpt from the opening remarks of Carroll D. Wright:
> It has been said that figures will not lie. It is equally true that
> liars will figure. It is our duty to prevent liars from figuring in
> the interest of any theory, by presenting original data fairly.
>
> http://books.google.com/books?id=YygLAAAAYAAJ&q=liars#v=snippet&
> JSTOR Stable URL: http://www.jstor.org/stable/2276665
>
>
> Newspaper coverage in Texas of the conference in 1889 did mention the
> remarks of Wright and did quote him delivering the maxim.
>
> Citation: 1889 July 12, San Antonio Daily Express, Page 2, Column 3,
> San Antonio, Texas. (NewspaperArchive)
>
> The national convention of the bureau of statistics of labor, which
> met in Hartford last month, had for its object the collection of
> facts, without reference to political or economical problems. … The
> president in his opening address says truly that "figures do not lie,"
> but he qualifies this by facetiously saying that "liars will figure."
> That is the great trouble with statistics; liars and cranks will add,
> subtract, divide and theorize.
>
>
> Another 1889 instance of the quote appears in an article arguing about
> sewer routes in California. The maxim is unattributed.
>
> Citation: 1889 August 5, Los Angeles Times, A Strong Argument: Against
> the Ballona Sewer Route, Page 5, Los Angeles, California. (ProQuest
> Historical Newspapers)
>
> Statements are easily made; their value, however, depends upon the
> reliability of the parties who make them. Figures don't lie, but liars
> will figure. I challenge an investigation of the situation
>
> http://news.google.com/archivesearch?q=%22liars+will%22+Ballona&
>
...
> Jumping forward over several years of citations to 1895, one finds
> that the attachment of the quote to Carroll D. Wright is strong.
>
> Citation: 1895 January, The Inland Printer, Changes Wrought by the
> Composing Machines by Hugh Wallace, Page 345, Maclean-Hunter Pub. Co.
> (Google Books full view)
>
> ... forces belief in Carroll D. Wright's saying that "figures don't
> lie, but liars will figure," ...
>
> http://books.google.com/books?id=3oHnAAAAMAAJ&q=liars#v=snippet&
>
>
> In 1898 a laudatory profile of Carroll D. Wright emphatically credits
> him with originating the maxim.
>
> Citation: 1898, Second Biennial Report of the Bureau of Labor of the
> State of New Hampshire: Volume IV, Labor Bureaus: Carroll Davidson
> Wright, page 89, Arthur E. Clarke: Public Printer, Manchester, New
> Hampshire. (Google Books full view)
>
> Carroll D. Wright, Commissioner of Labor, United States Department of
> Labor, Washington D.C ., who has just been honored with membership in
> the Institute of France, and honorary membership in the Imperial
> Russian Academy of Sciences, is one of the foremost statisticians of
> the world. … It was he who originated the now famous and much
> misquoted saying, "Figures do not lie, but liars figure."
>
> http://books.google.com/books?id=aS0oAAAAYAAJ&q=misquoted#v=snippet&
>
...
>
> While searching for predecessors I relaxed the requirement for the
> presence of the four words: figures, lie, liars, and figure.  A prolix
> conceptual antecedent in 1852 combines the phrase "figures won't lie"
> with the idea that individuals who use statistics are liars. The
> distrust of statistics has a long lineage that culminates in the
> phrase "lies, damned lies, and statistics".
>
> Citation: 1852 August, The New Englander, The Sources of our
> Population, Page 393, Volume 39, F. W. Northrop, New Haven.
>
> The old saying, that "figures won't lie," is true, without doubt; and
> the same may be said of letters, marks, and other signs of thought.
> But the mode in which many use figures, in order to carry a point, has
> sometimes tempted us to believe that the hasty remark of the Psalmist,
> if paraphrased thus - "all men" - who deal in statistics "are liars,"
> - is not far from the truth.
>
> http://books.google.com/books?id=cWdJAAAAMAAJ&q=liars#v=snippet&
>
>
> An antecedent in 1859 asserts that lies are inherent in figures and
> facts themselves. In this citation the liar with a "gifted
> imagination" is viewed positively. The quoted phrase in the cite below
> is a variant of an earlier phrase: "nothing is so fallacious as
> figures, except facts" that appears in 1846 or earlier.
>
> Citation: 1859, The Princeton Text Book in Rhetoric by M. B. Hope,
> Page 99, John T. Robinson, Princeton, New Jersey.
>
> Unreal pictures, by a gifted imagination, often give a truer
> impression in effect, than a literal description.
> It has been said, not less justly, than wittily, that "nothing lies
> like figures except facts."
>
> http://books.google.com/books?id=vVpZAAAAMAAJ&q=%22like+figures%22#v=snippet&
>
>
> An antecedent in 1869 moves closer to the form of the final maxim. It
> uses the phrase "figures won't lie" and the word "liars"; however, the
> liar is not externally identified. Rather, figures "are the greatest
> liars."
>
> Citation: 1869 November 20, The Week: United States, Page 2, Column 2,
> Halifax, Nova Scotia.
>
> The New York "World" has discovered, according to its own account, a
> deficit of over forty millions in his calculations. What a horrible
> satire it is to say that figures won't lie. They are the greatest
> liars in the modern world.
>
> http://news.google.com/archivesearch?q=%22horrible+satire+it%22&
>
> I hope these three examples give a flavor of the antecedents. The
> misleading nature of figures and statistics together with the
> deceptiveness and the people who employ them are a popular topic for
> maxims.
>
> Garson O'Toole
>

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org