Objectivity: True-to-Nature, Mechanical, and through Trained Judgment

The last post (Scientific Objectivity has a History) was about an article from 1992 called "The Image of Objectivity," by Lorraine Daston and Peter Galison.

Daston and Galison, 15 years later, have now written a book-length treatment of the topic, Objectivity (MIT Press, 2007). It argues that "To pursue objectivity--or truth-to-nature or trained judgment--is simultaneously to cultivate a distinctive scientific self wherein knowing and knower converge. Moreover, the very point at which they visibly converge is in the very act of seeing not as a separate individual but as a member of a particular scientific community."

Needless to say, given my comments in the prior post about this subject, I was very excited to see the book's arrival -- though I don't offer a review of the book here, only a notice. (If you're looking for a review, Ted Porter, historian of science at UCLA, has the first one out, in Nature [449, 985-987 (25 October 2007)].) From the MIT blurbers, we have this summary statement:

In Objectivity, Lorraine Daston and Peter Galison chart the emergence of objectivity in the mid-nineteenth-century sciences--and show how the concept differs from its alternatives, truth-to-nature and trained judgment. This is a story of lofty epistemic ideals fused with workaday practices in the making of scientific images.

As a taste, here are some images Daston and Galison use to show what they mean by the three kinds of representation in science, Truth-to-Nature, Mechanical Objectivity, and Trained Judgement.

The first is of truth-to-nature, which was drawn to portray "the underlying type of the plant species, rather than any individual specimen. It is an image of the characteristic, the essential, the universal, the typical..." (D and G, p. 20). (Also see related commentary here, in a post on epistemology for the Science Blogs Basic Concepts series.)

The next representation is a snowflake, meant to portray mechanical objectivity. It is shown "with all its peculiarities and asymmetries in an attempt to capture nature with as little human intervention as possible..." (p. 20).

The third one is meant to portray what they mean by Trained Judgment. It is an "image of the magnetic field of the sun [mixing] the output of sophisticated equipment with a 'subjective' smoothing of data--the authors deemed the intervention necessary to remove instrumental artifacts..." (p. 21).

This sequence fits with one of the more intriguing parts I recall from the earlier article version, which was the placement of photography within this history of objective image making. When we see the photograph as part of a much larger trajectory in ideals of objective representation, we get a sense of humility, I think, noticing that it is but one aspect of a bigger story. (Plus, the opening image of my prior post, of the camera obscura, shows that photography itself was a later innovation in degree, not kind.)

We might otherwise tend to assume the photograph is the end of the story, the most objective one could be. See, for example, the oft-used claim that 'a picture is worth a thousand words,' as with Tim at Deltoid's popular blog post, meant to show that the picture speaks truth, while it is the words that corrupt. Or, to put it in Daston and Galison's words, "What the photograph...offered was a path to truthful depiction of a different sort, one that led not by precision but by automation--by exclusion of the scientist's will from the field of discourse" (p. 117). But the photograph is not the end of the story, and for over a century scientists and others have had the chance to argue the point.

While reviewing three kinds of objectivity in their earlier article - Mechanical objectivity, Metaphysical objectivity, Aperspectival objectivity - Daston and Galison have this to say:

"The photograph that was the essence and emblem of mechanical objectivity carried no metaphysical cachet: at best it was an accurate rendering of sensory appearances, which are notoriously bad guides to the "really real." Nor would it have passed muster with the aperspectival ["the view from nowhere"] objectivity that eradicates all that is personal, idiosyncratic, perspectival. The photographic "look" was in fact radically perspectival--as many of our X-ray users never ceased to lament. We can fully understand why photographs wear the halo of objectivity only when we recognize that hid of objectivity that beatifies them is mechanical objectivity, and not its metaphysical or aperspectival kin..." (p. 123).

The book takes the story forward more, showing that the mechanical form of scientific objectivity was not the last, It was later supplanted by what Daston and Galison call trained judgment, a form that required human, not just mechanical, involvement in decisions. Reading a telescopic or microscopic or encephelographic or X-ray image requires training and judgment,. It requires an expert class, a group warranted as virtuous enough to do the job correctly. Daston and Galison do not argue that mechanical objectivity is gone, replaced by trained judgment,. They also do not argue that truth-to-nature has disappeared from scientific work. The latter does not replace the former, but stands as a new way to posit the relationship between observer and observed. But they do seem to suggest that "objectivity" was only one era in scientific image-making history, which has been followed by trained judgment. I think of it more in the sense that what we mean by objectivity has changed over time, not that it has come and gone. It's just that our sense of objective representation differs from earlier senses. In some cases, those differences are so stark that we don't even consider the earlier practices objective at all (as in truth-to-nature).

One little chart (p. 371) sums up the differences in epistemic virtues between the different ways of representation:

I'll leave the table for readers to consider, but wonder aloud what it means for practicing scientists when they confront the shifting virtues of science. Which will be next, for the next generation?

And so, without continuing into a full-on essay about the book alone, I leave this post as a notice of it. As intimated in the prior post, I also give notice of the new book here as a prologue to a series of posts over the coming weeks that might hold together the topics of objectivity, representation, truth, context, and proof.

For anyone who has kept up with the documentarian Errol Morris's fascinating series of mini-essays on the history of photography, I can admit here that this series was prodded along from his discussion. There will be much more to say about Morris's intriguing questions and cogitations on the subject, fear not.