As you might expect, I visit a lot of journal Web sites. I acquire cookies from many of these journal sites on my computer. So far, so normal.

Recently, however, I started to see ads from one particular journal following me across the Web, appearing in the oddest and most unexpected locations — on the New York Times, on the Atlantic, and on some other sites that have nothing to do with science, journals, or scholarship.

I was being followed!

From what I could deduce, DoubleClick — a Google property — was enabling this kind of ad following.

This is nothing new for the larger Web. Go to a travel site, and you’ll probably have hotel and airline ads following you soon enough. Go to a cooking site, and ingredients and restaurants will follow you. Go to a shelter site (e.g., Home Depot, Lowe’s), and suddenly faucets and garden tools are in your ads. Telemetry and triangulation are tracing tactics.

Traversing from general property to general property, these ads don’t leap out at you. They’re just kind of there, and could be there by chance or by design. It’s hard to know, and they register as normal or only slightly askew.

The challenge for scholarly publishers is that the zone of professional definition is potentially viewed as more sacrosanct than the zones of entertainment, vacation, hobby, or chore. Years of training, decades of reputation-building, and specialized knowledge sources all factor into a separate professional identity. To have markers from that identity popping up where you least expect it — well, it’s potentially jarring.

You may be a Red Sox fan and a cardiac surgeon in Boston, but do you really want a heart journal ad popping up on your fantasy baseball site? You might knit on nights you’re not coaxing lab rats along an experimental protocol, but do you want reagent ads popping up on your stitching sites?

Online advertising is a different beast. Facebook and Google are acquiring huge potential advantages in the consumer space because they know so much about us now. Their targeting capabilities are increasing daily. Leverage these completely, and they could shut the door for a decade.

Because online inventory is more finite than print inventory, the ability to advertise across sites will become increasingly important, especially for sites with smaller audiences and flat inventory (based on page views). But if scholarly publishers have perceptual problems advertising into non-scholarly properties — if my experience represents a common feeling others do or will share — their ability to expand their inventory by tracking users outside their sites ends. And that has revenue implications in the long run.

Professional identities among scientists often clash with personal identities. When advertising inadvertently drags the professional into the personal or vice-versa, the experience can be jarring. Because of that, scholarly publishers may have trouble following their users with ads.