Musings of a Philadelphia Physician who has served the community for six decades

Return to Home

On Writing History

History doesn't write itself.

More, or Fewer, Raisins in the Pudding

Some book publishers do indeed regard history as handfuls of paper in a manuscript package, mostly requiring rearrangement in order to be called a book. Librarians are more likely to see historical literature dividing into three layers of fact mingled with varying degrees of interpretation: starting with primary sources, which are documents allegedly describing pure facts. Scholars come into the library to pore over such documents and comment on them, usually to write scholarly books with commentary, called secondary sources . Unfortunately, many publishers reject anything which cannot be copyrighted or is otherwise unlikely to sell very well, however important historians say it may be. Authors of history who make a living on it tend to focus on the general reading public, generating tertiary sources, sometimes textbooks, sometimes "popularized" history in rising levels of distinction. Sometimes these authors go back to original sources, but much of their product is based on secondary sources, which are now much more reliable because of the influence of a German professor named Leopold von Ranke. Taken together, you have what it is traditional to say are the three levels of historical writings.

{Leopold von Ranke}
Leopold von Ranke

Leopold von Ranke formalized the system of documented (and footnoted) history about 1870, vastly improving the quality of history in circulation by insisting that nothing could be accepted as true unless based on primary documents. Von Ranke did in fact transform Nineteenth century history from an opinionated propaganda in which it had largely declined, into a renewed science. At its best it aspired to return to Thucydides, with footnotes. That is, clear powerful writing, ultimately based on the observations of those who were actually present at the time. Unfortunately, Ranke also encased historians in a priesthood, worshipping piles of documents largely inaccessible to the public, often discouraging anyone without a PhD. from hazarding an opinion. The extra cost of printing twenty or thirty pages of bibliography per book is now a cost which modern publishing can ill afford, making modern scholarship a heavier task for the average graduate student because the depth of scholarship tends to be measured by the number of citations, but incidentally "turning off" the public about history. All that seems quite unnecessary, since primary source links could be provided independently (and to everyone) on the Internet at negligible cost. And supplied not merely to the scholar, but to any interested reader, no need to labor through citations to get at documents in a locked archive.


The general history reader remains content with tertiary overviews, and a few brilliant secondary ones, because document fragility bars public access to primary papers; but many might enjoy reading primary sources if they were physically more available. The general reader also needs impartial lists of "suggested reading", instead of the "garlands of ibids", as one wit describes the bibliographies employed by scholars. If you glance through the annual reports of the Right Angle Club, you will see I have increasingly included separate internet links to both secondary as well as primary sources, because the bibliographies within the secondaries lead back to the primaries. Unfortunately, you must fire up a computer to access these treasures. The day soon approaches when scholars can carry a portable computer with two screens, one displaying the historian's commentary while the second screen displays related source documents. It seems likely history on paper will persist while that remains cheaper, but also because e-books make it hard to jump around. Newspapers and magazines particularly encounter this obstacle, because publication deadlines give them less time for artful re-arrangement. E-books are sweeping the field in books of fiction, because fiction is linear. Non-fiction wanders around, even though this subtlety is often unappreciated by computer designers.

The Prisoner in the Stone

{A Prisoner of the Stone}
A Prisoner of the Stone

The first revolution in book authorship has already taken place. Almost all manuscripts arrive on the publisher's desk as products of a home computer. In fact, many publishers refuse to accept manuscripts in any other form. Not only does this eliminate a significant typing cost to the publisher, but it allows him to experiment with type fonts and book design as part of the decision to agree to publish the book. What's more, anyone who remembers the heaps of paper strewn abut a typical editor's office knows what an improvement it can be, just to conquer the trash piles. To switch sides to the author's point of view, composing a book on a home computer has greatly facilitated the constant need for small revisions. An experienced author eventually learns to condense and revise the wording in his mind while he is still working on the first draft. However, even a novice author is now able to pause and select a more precise verb, eliminate repetition, tighten the prose. He can go ahead and type in the sloppy prose and then immediately improve it, word by word. The effect of this is to leave less for the copy editor to do, and to increase the likelihood the book will be accepted by the publisher. The author is more readily able to see how the book looks, than he would have been with mere type script.

It must be confessed, however, that a second revolution caused by the computer has already come -- and gone. Not so long ago, I once linked primary documents already on the Internet to the appropriate commentary within my blog. The idea was to let the reader flip back and forth to the primary sources if he liked, and without interrupting the flow of my supposedly elegant commentary. In those early days of enthusiasm, volunteers were eager to post source material into the ether, just asking for someone to read it eagerly. I had linked up Philadelphia Reflections to nearly a thousand citations when an invincible flaw exposed itself. Historians were eager enough to post source documents, but not so eager to maintain them. One link after another was dropped by its author, producing a broken link for everyone else. The Internet tried to locate something which wasn't there, slowing the postings to a pitiful speed. Reluctantly, I went through my web site, removing broken links and removing most links. Maybe linking was a good idea, but it didn't work. There is thus no choice but to look to institutional repositories for historical storage, and funding to pay for maintaining availability for linkage. It is not feasible to free-load, although only recently it had seemed to be.

This pratfall assumed technology would provide more short-cuts than in fact it would. A more difficult obstacle emerges after some thought given to the nature of writing history, because it seems remarkably similar to Michelangelo's description of how to carve a statue. Asked how to carve, he said it was simple. "Just chip away the stone you don't want, and throw it away." Michelangelo saw statues as "prisoners of the stone" from which it was carved, while insights and generalizations of history emerge from a huge mass of unsorted primary documents. But there is a different edge to writing history. Uncomfortably often, the process of writing history is one of disregarding documents which fail to support a certain conclusion, often documents which send inconvenient messages to modern politics. Carried too far, de-selecting disagreeable documents amounts to attempting to destroy alternative viewpoints. By this view of it, what the author chooses to disregard, is then as important as what he chooses to include. Unless awkward linkages are consciously maintained in some form, they will soon enough disappear by themselves. It is a great fallacy to assume that ancient history can be isolated from current politics, or even to believe that history teaches the present. Often, it is just the other way around.

Dithering History


Dithering is originally a photographic term, referring to the process of smoothing out rough parts of an excessively enlarged photo. Carried over to the profession of writing history, the term alludes to smoothing over the rough parts of a narrative with a little unacknowledged conjecturing.

In photography, when a picture is enlarged too far, it breaks apart into bits and pieces. Dithering fills in the blank gaps between "pixels" actually recorded by the camera. It amounts to guessing what a blank space should look like, based on what surrounds its edges. In the popular comic strip, Blondie's husband Dagwood works for an explosive boss called Mr. Dithers, who "dithers" between outbursts, when he vents his frustration. But that's to dither in the Fifteenth century non-technical sense, meaning to hesitate in an aimless trembling way. If the cartoonist who draws Blondie will forgive me, I have just dithered in the sense I am discussing in history. I haven't the faintest idea what the cartoonist intended by giving his character the name of Mr. Dithers, so I invented a plausible theory out of what I do know. That's what I mean by a third new meaning for dither, in the sense of "plausible but wholly invented". It's fairly common practice, and it is one of the things which Leopold von Ranke's insistence on historical documentation has greatly reduced.

To digress briefly about photographic dithering, the undithered product is of degraded quality to begin with. Dithering systematically removes the jagged digital edge, and restores the original analog signal. In that case, dithering the result brings it back toward the original picture. That's not cheating, it's the removal of a flaw.
Color Dithering

Dithering of history comes closer to resembling a different photographic process, which samples the good neighboring pixels on all sides of a hole and synthesizes an average of them to cover the hole. It works best with four-color graphics, converting them to 256-color approximations. But the key to all photographic approaches is to apply the same formula to all pixel holes. That's something a computer can do but a historian can't, and historical touch-ups seem less legitimate because the reader can't recognize when it has happened, can't reverse it by adding or removing a filter. Dithering may somewhat improve the readability of the history product occasionally, usually at the expense of inaccuracy, sometimes large inaccuracy. It's conventional among more conscientious writers to signal what has happened by signaling, "As we can readily imagine Aaron Burr saying to Alexander Hamilton just before he shot him." That's less misleading than just saying "Burr shouted at Hamilton, just before he blew his brains out." The latter sends no signal or footnote to the reader, except perhaps to one who knows that Hamilton was shot in the pelvis, not the head. In small doses, dithering may harmlessly smooth out a narrative. But there's a better approach, to write history in short blogs of what is provable, later assembling the blogs like beads in a necklace. Bridges may well need to be added to smooth out the lumps, but that becomes a late editorial step, consciously applied with care. And consequently, author commentary is more likely to recognized as commentary, rather than rejected as fiction.

Dithering the holes is often just padding. It would be better to spend the time doing more research.

SMTP Authorization and Handling Bounced Emails with PEAR Mail

Recently our ISP started requiring user signon in order to send emails. PHP's mail function stopped working as a result.

Naturally, the ISP did not notify us of this change so we were quite surprised when many thousands of emails on our newsletter list were rejected (every one of them, in fact).

What error message was returned to us to notify us of what the problem was? Why this helpful note:

Mail sent by user nobody being discarded due to sender restrictions in WHM->Tweak Settings

Doesn't that just say it all?

I'm being snide, but our ISP is really quite good about keeping its software up to date and aside from an occasional surprise like this, they are very reliable. Being up to date included the automatic incorporation of the PEAR Mail facility which we are now using.

PEAR's Mail system works quite well but two problems were very vexing until we stumbled our way to a solution:

  1. How, exactly, do we sign on to the SMTP server?
  2. How do we ensure that bounced emails (the bane of all email lists) get returned to us?

You might not think that the first question would be so hard but it actually took a good deal of trial and error to get it right. As for the second question, there is an awful lot of wrong information available out in Internet land (including but not limited to VERP and XVERP which I advise you to avoid).

With PEAR Mail you first set up a "factory" and then send emails, either singly or in a loop. We keep the user id, password, etc. in a file "above" the web server in hopes that will keep them secret ... here's the code (it actually is in production and it does in fact work):


# the email constants are contained in a file outside the web server

$headers = array (
         'From' => '"name"<>',
         'Sender' => '"name"<>',
         'Reply-To' => '"name"<>',
         'Return-Path' => '',
         'Content-type' => 'text/html; charset=iso-8859-1',
         'X-Mailer' => 'PHP/' . phpversion(),
         'Date' => date("D, j M Y H:i:s O",time()),
         'Content-Language' => 'en-us',
         'MIME-Version' => '1.0'

// call the PEAR mail "factory"
$smtp = Mail::factory('smtp',
      array (
            'host' => EMAIL_HOST,
            'port' => EMAIL_PORT,
            'auth' => true,
            'username' => EMAIL_USERNAME,
            'password' => EMAIL_PASSWORD,
            'persist' => true,
            'debug' => false
            ), '-f'

# to send emails:
# $headers['To']      = $to;        # provide the "$to" variable, something like $to = '"name"<>';
#                                   # note that the first parameter of $smtp->send can be "decorated" this way or just a naked email address
# $headers['Subject'] = $subject;   # provide the "$subject" variable
# $mail = $smtp->send($to, $headers, $contents_of_the_email);
#                          -------- ................................> except for 'To' and 'Subject',
#                                                                     $headers is provided by this module but can be over-ridden
# if (PEAR::isError($mail))
# {
#   echo "<p style='color:red;'>The email failed; debug information follows:<br />";
#   echo $mail->getDebugInfo() . "<br />";
#   echo $mail->getMessage()   . "</p>";
# }
# else
# {
#   echo "<p>email successfully sent</p>";
# }


My thanks to for the HTML entites conversion.

Please Let Us Know What You Think

(HTML tags provide better formatting)

Because of robot spam we ask you to confirm your comment: we will send you an email containing a link to click. We apologize for this inconvenience but this ensures the quality of the comments. (Your email will not be displayed.)
Thank you.