I greatly enjoyed reading David Mermin's last two
Reference Frame columns on factoring and quantum computing (PHYSICS TODAY, April 2007, page 8;
October 2007, page 10). However, as the proud one-time owner of the California license plate QUBITS
(which I had occasion to park beside a QUARKS license plate from New Mexico), I feel I must cheerfully
disagree with Mermin's choice to denote quantum bits as qbits. His slander of the traditional shortened
spelling of quantum bit as "the vulgar spelling qubit" lays down the gauntlet to the proud
worldwide community of heterographic homophone lovers, many of whom I call friends.
Mermin's arguments are
deceptively enticing:1 that qubit violates the English rule that qu
should be followed by a vowel, that no one would ever call the swab a Qutip, and that Paul Dirac with
good reason called them q-numbers, not qunumbers. One reason might have been enough, but cube it
and certainly no one will argue back at you.
Let us rebut those arguments
one by one. First, if Mermin is serious about his respelling of the shortened form of quantum bits,
then certainly he should choose q-bit instead of qbit, because the only word with a consonant following
a q in the Oxford English Dictionary is the crippling child of circumstance, qwerty,
to describe a keyboard. And, having settled on q-bit, one faces a dilemma. Writing "q-bit" immediately
conjures up Q-tip and, worse for those of us from the video game age, Q-Bert (or Q*Bert), the name
of an addictive video game featuring an eponymous alien who hopped on a tricolored pyramidal staircase.
Mermin may find no discomfort in hearing echoes of "Q-tip," but I think the majority of physicists—not
to mention computer scientists—would prefer not to be reminded of either ear infections
or hopping aliens when they transcribe their serious scientific work.
I note that q-number and
Q-tip arose at the same time. (Originally called "Baby Gays," Q-tips were renamed in 1926. The Q
stood for quality.) With no disrespect to Dirac, who seemed to have a thing for full-letter initials,
history, as measured by the success of q-number versus the ubiquity of Q-tip, has
clearly decided that the q- prefix works better for brand names than for scientific terms.
In short: Let us keep the corporate branding of science off our beloved quantum bits.
Most importantly for the
progress of science, the fact that qubit is an intentional homophone of cubit is
not vulgar but rather a blessing in disguise. A cubit is an ancient unit of length, roughly equal
to the length of a forearm. And that "roughly" should offend the heart of any physicist. Like the
foot, the cubit is an imprecise unit whose lack of precision makes me think it deserves to die. Can
you use cubit on your physics tests? No! And I can think of no better death than its replacement
by qubit, a precise unit of quantum information.
As an addendum to the beauty
of this homophone, one should not overlook the fact that the cubit is most often associated with
lengths in biblical texts. And despite Mermin's claim to the contrary, qu followed by a
consonant does appear in English words, albeit proper nouns, notably Qumran, the name of the settlement
nearest where the Dead Sea scrolls were discovered.
Long live qubits. And may
qbits and their dirty cousins Q-bits only rear their heads in a future quantum computing company
or in a medical journal describing bits of cotton swab lodged in the ear.
Mermin replies:
The enticing arguments that Dave Bacon tries so well to refute were not in my Reference Frame columns,
which I'm delighted he enjoyed. The point is put best in my new book,1 where I also note
that q-bit would indeed be better than Qbit, were it not that common expressions
such as 2-Qbit gate fail to work when written in the form 2-q-bit gate. To solve the
problem, I drop Dirac's hyphen but put Q in uppercase (which the qbit Bacon wrongly attributes
to me does not) to emphasize that it functions as the name of a letter and not as part of another grotesquely
spelled word.
Qbit for the quantum
object parallels the equally important Cbit for the classical object (clbit would be ridiculous).
The capitalization convention gives rise to the useful term Bit, which refers collectively
to both kinds of physical objects, as opposed to bit, which refers only to a number that can
be 0 or 1.
I agree with Bacon that
qubit has achieved its current status because of its spurious resemblance to cubit,2
but I do not agree that the foot deserves to die. It should be redefined 5 millimeters downward to
a light nanosecond3 a couple of years after our backward nation finally goes metric,
which I confidently expect to happen in a century or two.
Setting aside the fact
that transliterations of foreign proper nouns notoriously violate all kinds of rules, I note that
Qumran is pronounced k'mran. This class of examples would require qubit
to be pronounced k'bit.
It may be quixotic (but
certainly not Qxotic) to try to correct the spelling of an entire community, but I owe it
my best shot. What else is retirement good for?
References
1. N. D. Mermin, Quantum Computer Science: An Introduction, Cambridge U. Press, New York (2007), p. 3.
2. Ref. 1, p. 4.
3. N. D. Mermin, It's About Time: Understanding Einstein's Relativity, Princeton U. Press, Princeton, NJ (2005), p. 22.
Featured Jobs
