Features
Part of Why commencement speeches face the worst listening conditions
Commencement speech misquotes: five patterns and six checks
Commencement speeches misquotes explained through five patterns of invention and six checks in order, with the Vonnegut-MIT sunscreen hoax as the named case.
Commencement speeches misquotes follow a failure mode all of their own: the graduation address that was never given. A column, an essay, a satirical piece or an anonymous list of advice circulates with a university's name attached, acquires a speaker, and is then quoted for years as a speech.
This page walks through five patterns of invention and six checks in order, using named cases where the record allows, and closes with how to correct one without amplifying it.
What to take away
- Before checking the wording, check that the speech happened. In this genre that is not a formality.
- Circulated graduation texts acquire a speaker and an institution after the fact, and both are usually wrong.
- The institution's registrar or ceremony office can settle the question in one email, and almost nobody asks.
Why this genre attracts invention
A graduation address is advice to young people from someone older, delivered at a moment of transition. That is an appealing frame for any piece of writing, and attaching it to a speech gives the advice an authority that a column does not have.
The text also travels well. It is self-contained, it needs no context, it is about the right length to forward, and it arrives with a ready-made scene. The mechanics are the ones described in accounts of the chain letter: a text that carries its own reason to be passed on will outrun any correction attached to it.
The five patterns
- The column promoted to a speech. A newspaper piece circulates without its masthead, gains a stage, and eventually gains a speaker. The textbook case: Mary Schmich's satirical commencement column for the Chicago Tribune exploded into an Internet hoax disguised as an address supposedly delivered at MIT's 1997 Commencement by author Kurt Vonnegut, the so-called "sunscreen speech."
Five Patterns of Invention
Pattern
- Column promoted
- Gains a stage
- Wrong institution
- Moved to famous school
- Wrong speaker
- Reattributed to famous name
- Composite
- Several addresses stitched
- Improved version
- Edited until sharper
What Changes
- Column promoted
- Gains a speaker
- Wrong institution
- Wrong speaker
- Composite
- Improved version
Checking, in order
- Did the ceremony happen and who spoke? Ask the institution. Registrar, ceremony office or archive. Search the Internet Archive (archive.org) for the earliest dated appearance of the wording. This step alone resolves most cases.
Checking a Suspect Speech
- Ask institutiondid ceremony happen, who spoke
- Find the program naming speaker and date
- Look for institutional recording or published text
- Date the earliest appearance of the wording
- Check whether campus paper reported it
- Identify the text's actual author
The standard methods of fact-checking apply here without modification, and the reason they work so well is that the underlying question is unusually binary: either an institution held a ceremony at which a named person spoke, or it did not.
The reverse case
The opposite error is also common: a real address dismissed as fabricated
Genuine or Invented
Signal
- No institution in early versions
- Invention or reattribution
- Named in institution's program
- Genuine
- Identical across sources
- A printed original
- Varies between sources
- Oral or edited
- Campus paper covered it
- Genuine
- Only aggregators carry it
- Unverified
Points Toward
- No institution in early versions
- Named in institution's program
- Identical across sources
- Varies between sources
- Campus paper covered it
- Only aggregators carry it
| Signal | Points toward |
|---|---|
| No institution named anywhere in early versions | Invention or reattribution |
| Institution names it in its own program | Genuine |
| Text identical across sources, word for word | A printed original somewhere |
| Text varies substantially between sources | Oral circulation, or successive editing |
| Campus paper covered it | Genuine |
| Only national aggregators carry it | Unverified |
Correcting one
Lead with what the record shows, name the institution you contacted, and report what it said. State which of the four claims failed: the wording, the speaker, the occasion, or whether a speech occurred at all.
Do not reprint the disputed text to correct it. Reprinting spread it, and if the text is somebody's copyrighted column, reprinting causes trouble. Status questions are in public domain speeches.
Where you can identify the original author, say so and credit them, which is usually the correction people want. It turns a debunking into an attribution, which is the more useful outcome.
General handling of disputed wording in ceremonial material is covered in award speeches. The version-history habit that catches most of these errors early runs through political speeches.
Genuine addresses, ordinary errors
Once a speech is confirmed, the usual problems return. The institution's published text may be the prepared version, and the recording may be edited for length.
A quotation may come from the question period, not the address, and the working method is in commencement speeches.
Common questions
A university has no record of a speech everyone remembers. What does that mean?
Usually that the speech did not happen there. Occasionally that the archive is incomplete. Ask which, and report the answer you were given.
Is it worth correcting a text that has circulated for years?
Yes, and expect it to keep circulating. A short, findable correction that names the real author does more good than an argument.
The institution published a text that differs from the video. Which is right?
The video is what happened. The published text is what the institution decided to publish, sometimes with the speaker's edits. Report both.
Can I quote a line to identify which version I mean?
A short identifying phrase is usually unavoidable in a correction, but it is rarely necessary to reproduce more than that, and describing the passage keeps you clear of a permissions question entirely.
