Real Spellers

English Makes Sense!

Got a question for y'all (hey, I live in NC now, so I'm allowed to say that now!) I have a student exploring the < spire > family, and she can't figure out why all the words she has found in that family retain the < s > from the Latin spirare, except for < expire >.  I know the < s > isn't needed phonologically, but it doesn't interfere with it either, and I thought retaining the etymological link was more important. I notice in the "70 Matrices" that the < ex- > prefix is left out. Any thoughts? 

Comments (2)

This comment was minimized by the moderator on the site

From Pete:

Good one!

This is a rich common question that comes up and I know that there are other places this issue of an prefix before a base or stem beginning with has a similar process.

Gail and/or Rebecca is this a topic you are well versed in? I can respond, but it would require some research etc. I’m pretty sure I’ve even heard the name for the process that is going on here, but can’t remember off the top of my head. If either of you already have a good sense of how you would respond, please do!

Glad you’re doing work that provokes such a rich question Matt!

More soon,


Comment was last edited about 1 year ago by Matt Matt
This comment was minimized by the moderator on the site

From Gail:

I'm glad you asked this question, because while I've thought about this many times before, I'm looking at it more deeply.

Here are my thoughts:

There are indeed many other places where this comes up, the most common ones in my experience being <expire> and <expect>. The Romans spelled these words both ways.

Here is how they are listed in my Oxford Latin:

ex(s)to (my favorite because I never noticed, or didn't remember, the <st> base in <extant>)!
ex(s) tinguo

Lewis and Short list both forms but separately; e.g., exsto and exto, exstinguo and extinguo. We don't have either <exstant> or <exstinct> in English, but <exstant> has been attested in the past.

Douglas seems to include both roots - for as many of these words as I've looked at. I suspect that all or most of the <ex> words that are followed by <s> had this double life in Latin.

Out of curiosity, I looked up words with <xs> on word searcher and all that comes up on the regular search is <coxswain>, which is particularly interesting since its pronunciation is so different from its spelling - you might think if anywhere a change in spelling might occur it would be in that word. But the free BSD turns up many words spelled with <exs> though not <exspire> or <exspect>.

Interestingly, ithe free BSD includes <exscind), which might seem particularly redundant in English, where the /s/ seems to be triply represented in the spelling (<x, s, c> but not so much in Latin where the <c> represents /k/. Perhaps this survives because we are familiar with <sc> as a digraph. The OED calls the spelling <excind> "erroneous"!

I followed up in the OED a few <exs> words that came up in Word Searcher, I came across <exsert>, used in biology to mean "to thrust forward or protrude" : "The body is exserted through the brachial slit." But the spelling <exert> for that denotation is extinct. And if you look at the different denotations of <exert>, they are all labeled as obsolete except the sense we know best.

This surprised me:

Etymology: < Latin exert-, better exsert- , participial stem of exserĕre to put forth, bring out, < ex- out + serĕre to bind, entwine. The formation is probably due to antithesis with inserĕre to insert v. See exsert v

("Antithesis" is a new linguistic term for me.)

Matt, you have expressed surprise that English hasn't retained the <s> in the spelling for etymological reasons. We certainly do notice many instances in our orthography where letters have survived in spellings particularly for that reason. But we also find instances (like <expire> and <expect>) where our orthography has not held onto a spelling to preserve an etymological link. It seems that both types of evolution occur. And why not?

In the course of this investigation, I have come across a word that is new to me - one that behaves more in accordance to your expectations:


Etymology: < classical Latin discinctus not wearing a belt or girdle, wearing loose clothes, undisciplined, easy-going, use as adjective of past participle of discingere to remove the belt or girdle from < dis- dis- prefix + cingere to gird (see cinct adj.).

Now rare.
Lax, undisciplined; dissolute. Later also (with conscious reference to the etymology): not wearing a belt or girdle.

The presence of the <c> in <discinct> makes sense in terms of the root. So this word follows your expectation of retention of a letter for etymological reasons. The words <discern> and <discept> do, too.

Why has <exsanguine> prevailed in English and not <exanguine>, <exsiccant> but not <exiccant> and yet both <exstrophy> and <extrophy>survive?

I think this is an area of interesting variation in our spelling system.

As to constructing matrices that do not include <expire> and <expect> in the families of <spire> and <spect>, this has long been my practice. I leave them off my matrices and use them as a context for this very interesting discussion!

I would love to know whether the Roman grammarians had anything to say about these spellings with and without the <s>. Possibly some writers used the <s> and others didn't. Either way, their readers would have known what they meant.


Comment was last edited about 1 year ago by Matt Matt
There are no comments posted here yet