Further adventures with Low Attachment

Bonnie Taylor-Blake to ADS-L on 8/10 under the heading “Another zoological crash blossom”:

The headline for a blog post hosted by the Smithsonian:

“Scientists track a mysterious songbird using tiny backpack locators

This reminded me of a favorite from a few years ago, “Public urged to keep track of squirrels with mobiles.” (See Ben Zimmer’s column about this and other crash blossoms [here].)

Two ambiguous headlines that might be understood in an unintended way because of how modifying phrases (underlined above) are attached to preceding material:

parsed low in constituent structure, by attaching to a small preceding constituent (using tiny backpack locators attached to a mysterious songbirdwith mobiles attached to squirrels);

or parsed higher in constituent structure, by attaching to a larger preceding constitent (using tiny mobiles attached to track a mysterious songbirdwith mobiles attached to keep track of squirrels).

In both cases, high attachment (HA) was intended, but low attachment (LA) is sufficiently alluring as a default parsing strategy that at least some people will entertain the preposterous interpretations in which a songbird uses tiny backback locators and in which squirrels have mobiles. (See my 6/28/12 posting “The lure of Low Attachment”.)

Ovenbird equipped with a tiny GPS tracker (Smithsonian Magazine)

Attachment ambiguities are scarcely limited to headlines — in a little while, I’ll discuss a non-headline example from Wikipedia text; for more examples, check out the postings summarized in my Page on attachment — but they’re especially likely to occur there, for (at least) two reasons:

Headlines are compressed, abbreviated, in such a way that many cues to interpretation are eliminated; in the examples above, HA could be cued through by using (by using … locatorsby using mobiles) rather than just using and with, but at the cost of extra words (and if you actually wanted LA, you could cue it by using full relative clauses rather than reduced constructions: … songbird that uses … locators; squirrels that have mobiles).

Headlines rarely have preceding linguistic context that could bias the reader towards one parsing or the other, so the reader is obliged to treat them as if they had just dropped out of the sky.

Readers do have some expectations that can guide their interpretation, though these often can work at cross purposes and can conflict with a preference for LA:

It’s routine to organize sentences as having a topic followed by a comment on that topic.

Sentence topics tend to be discourse topics as well, so that a neutral way to organize a sentence is with old information (in some sense given in the context) before new.

The default for the arrangement of material in newspaper stories is to put the most important information first.

In the two examples above, the placement of the modifiers using … locators and with mobiles at the end of the headlines treats the information in them, correctly, as comment and new — but at the cost of potential ambiguity.

The ambiguity could be avoided by making the modifiers clause-initial (because the information they convey is important to the story), but at the cost of treating this information as if it were old:

Using tiny backback locators, scientists track a mysterious songbird

Public urged: use mobiles to keep track of squirrels

The first of these unfortunately introduces the possibility that it’s the scientists, rather than the songbird, wearing tiny backpack locators. Things can be clarified further with more words:

To track a mysterious songbird, scientists strap tiny backpack locators on it

Scientists strap tiny backpack locators on a mysterious songbird to track it

There are other possibilities, of course. (And potential ambiguities haven’t been eliminated: for instance, a mysterious songbird could refer either to a particular individual or to a type of songbird.)

Beyond headlines. From yesterday’s posting “Swiss art supplies in the morning”, this Wikipedia quote:

[graphite] stems from graphein, meaning to write or draw in Ancient Greek.

LA: in Ancient Greek attached to write or draw. HA: in Ancient Greek attached to meaning to write or draw.

As with many attachment ambiguities, I’m unsure about whether the ambiguity is sufficiently troublesome to readers to call for mockery and rewriting. This one strikes me as problematic only for people who are looking for trouble. But I could be wrong.

One Response to “Further adventures with Low Attachment”

  1. Robert Coren Says:

    This one strikes me as problematic only for people who are looking for trouble.

    You know me so well. But I will insist that the low attachment gave me a moment’s pause, even though I knew perfectly well what was meant.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: