Attachment ambiguity in "Frazz"

« previous post | next post »

Today's "Frazz" (via Ed Cormany on Twitter):

The confusion here is reminiscent of the discussion on Neal Whitman's Literal-Minded blog about the expression, "I need to pee like a racehorse." Neal explains, "This is an example of an attachment ambiguity, in that we could theoretically attach the modifier like a racehorse to the 'lower' verb phrase pee or to the 'higher' verb phrase need to pee." Thus, it's not:

I [ need to pee ] [ like a racehorse ]

but rather:

I [ need to [ pee like a racehorse ] ]

In the "Frazz" strip, the ambiguity works the other way. The teacher thinks like my dad attaches to the lower verb be:

I [ want to [ be the guitarist for Iggy and the Stooges like my dad ] ]

when what's intended is an attachment to the higher verb want:

I [ want to be the guitarist for Iggy and the Stooges ] [ like my dad ]

See Neal's blog for more examples of attachment ambiguity, many helpfully parsed with syntactic tree diagrams. And see also Arnold Zwicky on "the lure of Low Attachment" here (with links back to relevant LL posts).


  1. Neal Goldfarb said,

    April 12, 2013 @ 9:41 am

    It's also an ellipsis ambiguity: "like my dad (does)" vs. "like my dad (is)".

  2. S. Norman said,

    April 12, 2013 @ 10:09 am

    I wouldn't wann be Ron Ashton. He's dead.

  3. Michael Watts said,

    April 12, 2013 @ 10:14 am

    > It's also an ellipsis ambiguity: "like my dad (does)" vs. "like my dad (is)".

    I don't understand what's being claimed here… the choice between 'like my dad does' and 'like my dad is' isn't an additional ambiguity; it's coincident with the parse structure: [I [want to [be a guitarist like my dad is]]] vs [I [want to be a guitarist] [like my dad does]]. Would you also say that, in addition to being ambiguous between ellipsed "like my dad (does)" and "like my dad (is)", it's also ambiguous between "like my dad (wants to)" and "like my dad (is)"?

    Compare "I want to climb mountains like my dad". It has exactly the same structural ambiguity, and no "ellipsis ambiguity": [I [want to [climb mountains like my dad (does)]]] vs [I [want to climb mountains] [like my dad (does)]]. But the only difference here is that 'want' and 'climb' both require do-support, whereas 'be' is auxiliary in itself.

  4. Jason Merchant said,

    April 12, 2013 @ 10:19 am

    And the ambiguity is found in comparatives as well: "I want to be richer than you" can mean "richer than you are" or "richer than you want to be" (ellipsis can preserve the ambiguity or not, depending on the nature of the predicates involved, as Neal points out). And what's cool is that some languages have morphological means for indicating the low attachment: modern Greek, for example, uses the genitive on "you" to force the low attachment (while also having a PP "than you" which can be ambiguous: see here for more than you probably wanted to know).

  5. J.W. Brewer said,

    April 12, 2013 @ 10:28 am

    Ron Asheton is dead, but now James Williamson is back in the band. I would say that "being guitarist for Iggy and the Stooges" != "being Iggy Pop's guitarist," but compared to the bogus prescriptivist ideas the cartoon teacher probably has about English grammar that's a pretty picayune issue.

  6. Eric P Smith said,

    April 12, 2013 @ 10:30 am

    > It's also an ellipsis ambiguity: "like my dad (does)" vs. "like my dad (is)".

    I'm with Neil. There is only one ambiguity, but it lies in two dimensions, attachment and ellipsis. The two possible meanings differ both in attachment and in ellipsis.

  7. Maryellen MacDonald said,

    April 12, 2013 @ 11:22 am

    For information about how the statistics of English make low attachment the dominant interpretation, see the article below. The story basically goes like this:
    1. There are alternative structures to convey the high attachment meaning without using the V…V…modifier structure, such as, "Like my dad, I want to be the guitarist for Iggy and the Stooges." These alternatives are preferred by English speakers for a variety of reasons (preferences for end-weight, among others).

    2. As a result of these preferences for alternative forms to convey the meaning, utterances with an intended high attachment interpretation are rare; corpus analyses show that low attachment is overwhelmingly the dominant pattern in English.
    3. Perceivers (like the teacher in the Frazz strip) have implicitly learned these language statistics and are guided by them in ambiguity resolution, so that they adopt the dominant low attachment interpretation of these ambiguous structures.

    MacDonald, M. C., & Thornton, R. (2009). When language comprehension reflects production constraints: Resolving ambiguities with the help of past experience. Memory & Cognition, 37, 1177-1186.
    downloadable here:

  8. Boris said,

    April 12, 2013 @ 2:23 pm

    I [ want to be the guitarist for Iggy and the Stooges ] [ like my dad ] doesn't quite mean to me what it needs to in order to make this work. I understand the above as "I want to be a guitarist…in the same way my dad wants to be a guitarist…" which makes no sense. Are there multiple ways of wanting to be a guitarist for Iggy Pop? And even if there are, they'd only be of interest to psychologists and the like.

  9. Faldone said,

    April 12, 2013 @ 2:35 pm

    I don't see this as comparing multiple ways of wanting to be guitarist for Iggy Pop but rather comparing wanting to be guitarist for Iggy Pop vs. not wanting to be guitarist for Iggy Pop. There might well be multiple ways of not wanting to be guitarist for Iggy Pop, but that's not exactly the point here.

  10. Brett said,

    April 12, 2013 @ 4:40 pm

    I had an altogether different reading:

    I [ want to [ be the guitarist for Iggy and [ the Stooges like my dad ] ] ]

    This hasn't got much going for it, pragmatically, but that was my initial parsing. I suspect that, having never been a fan and being too young to hear about Iggy Pop at the height of his popularity, "Iggy and the Stooges" doesn't feel like a fixed phrase for me.

  11. Neal Whitman said,

    April 12, 2013 @ 4:53 pm

    Brett: As I wrote to Ben..

    Of course, "like my dad" could attach lower still, as an adjective phrase modifying "stooges". But since "Iggy and the Stooges Like My Dad" isn't a known band, both of the other parses are easier to get.

    Except, as you (Brett) point out, for people for whom the band name is unfamiliar.

  12. Christopher Hodge said,

    April 12, 2013 @ 6:18 pm

    Gosh, am I the only person who parsed this as two declaratives, viz [I want to be the guitarist for Iggy] and [the Stooges like my Dad]?

  13. The Ridger said,

    April 12, 2013 @ 6:20 pm

    @Boris: you're misunderstanding "like", which doesn't mean the boy wants to be the guitarist *in the same manner of wanting* as his father, but only that, *as his father does*, he wants to be the guitarist.

  14. Pseudonym said,

    April 12, 2013 @ 8:39 pm

    The actual guitarist for The Stooges is a VP at Sony. He has officially sold out and become Record Company Scum. Definitely not someone you want to emulate.

  15. Chris C. said,

    April 12, 2013 @ 10:01 pm

    the bogus prescriptivist ideas the cartoon teacher probably has about English grammar

    That'd be Mrs. Olsen. Miss Plainwell here is pretty cool.

  16. Nathan Myers said,

    April 14, 2013 @ 12:35 am

    Can't everyone reading here quote the lyrics from the Monty Python troupe's Lumberjack Song?

    "… I wish I'd been a girly, just like my dear papa."

    I only just noticed that this offers a third alternative, that his papa wished his son had been a girly.

  17. Gene Callahan said,

    April 14, 2013 @ 11:46 am

    Well, Boris, that is a weird and unnatural way to read that.

  18. un malpaso said,

    April 14, 2013 @ 6:12 pm

    @Nathan: Wow. That just blew my mind. My first assumption was always "Papa was a girly." I was vaguely aware of the second alternative. But I never thought of the third one. Now, I wonder how many other syntactic constructions have layers of possible meanings that I never even considered?

    It seems to me, though, that a good rule of thumb is the deeper you have to dig for an interpretation, the less likely it was the original intended meaning. Still… all three could be true at the same time!

  19. Jason Merchant said,

    April 15, 2013 @ 9:49 am

    @Nathan: You've discovered Dahl's puzzle (aka one of the "eliminative puzzles of ellipsis", as named and discussed in Fiengo and May's 1994 book)…

  20. Rod Johnson said,

    April 17, 2013 @ 10:17 am

    Not precisely the same phenomenon but near enough that it merits mention here… I was amused by this bit of ambiguity herein: "First of all, I’m in a long-term committed relationship with a child…"

RSS feed for comments on this post