Arnold Zwicky's Blog

From Nicholas Kristof’s NYT opinion column 1/8 (“A Poverty Solution That Starts With a Hug”):

The science is still accumulating. But a compelling message from biology is that if we want to chip away at poverty and improve educational and health outcomes, we have to start earlier. For many children, damage has been suffered before the first day of school.

Then on 1/9 commenter Mar Rojo (on Barbara Partee’s 1/4 Language Log posting “Nate Silver knows his passives”) added a query about that last sentence, and e-mailed the query to me:

After reading Geoff Pullum’s description of the passive in English, I thought I knew my passive. Now I’m not sure. A few commenters on a certain language forum have claimed this as passive: “For many children, damage has been suffered before the first day of school.” Is it?

The short answer is Yes.

Now for the details:

Remove the initial sentence adverbial for many children, and you have:

(1) damage has been suffered before the first day of school

This is an agentless passive clause; the corresponding active is:

(2) X has suffered damage before the first day of school [identity of X not specified]

Compare an agentive passive and its active counterpart:

(1′) damage has been suffered by many children before the first day of school

(2′) many children have suffered damage before the first day of school

I can see two reasons why this analysis might not be obvious to everyone. One possible reason is that (2) and (2′) have subjects that don’t denote Agents and verbs that don’t denote Actions; instead, the subjects denote Experiencers and the verbs denote Experiences. Some discussion of participant roles of subjects here:

Despite this fact, such active clauses have perfectly ordinary passive counterparts (with the direct object in the active clause — here, damage — corresponding to the subject in the passive clause). This is a problem only if you subscribe to the Agent-Action (semantically based) view of clause types; see my posting linked to above.

Another possible reason is that in the original example from Kristof, the subject of the corresponding active is expressed in that initial sentence adverbial attached to the passive clause, rather than in a VP-adverbial with the preposition by (as in (1′)). So the original example doesn’t quite look like a canonical passive. But it’s an allowable variant.

This entry was posted on January 15, 2012 at 10:13 am and is filed under Passive, Syntax. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.