Collections, connections, corrections, and convection: - CUM dancing

More tales from the word front

Thinking, as I have been, about sonorants, I was more than usually intrigued by a friend's question about words spelt in English either com- or con- (and, I have since realized, two other sonorants come into the same story: coll- and corr-).

Many such words derive in some way from a Latin word that uses CUM- as a prefix. But there are various ways that the M develops. In most cases the CUM- becomes con- but there are others. In this back-of-an-envelope table I give examples of the main ways:

The sort of Latin used to produce Romance vernaculars ironed out irregularities (eg "sing": cano canere cecini cantum was hard to learn for speakers of Latin for Speakers of Other Languages (LSOL – which is after all what we‘re dealing with); so those speakers preferred the regular canto cantare cantavi cantatum. Which gives  Italian, ‘cantare‘  Sp, Pg, Catalan cantar‘, Fr chanter etc etc.

As a result, generally, irregular verbs don't fare well in forming vernacular words (apart from learned/technical words like conference). The Italian for CARRY is nothing to do with ferre; it's sopportare, portare {Latin}] was easier to handle; the prefix in that case [I used Italian as an example, because that was my interlocutrix‘s  {not sure if that‘s a word; but it is now} focus] is sub).

This isn't to say that ferre has left no trace in modern Italian; but those traces are not obvious and well hidden. The effective parts of the verb (the ones people learn...

<autobiographical_note  essentiality="0"=>
The first time I met the word paradigm it was in a Greek Grammar book: if you learn the standard model (not that one, SILLY) you can work out any part of a verb on the basis of those four parts.
</autobiographical_note>

... to work out all parts of the verb) are fero, ferre, tuli, latum. (Compare this with 'to love': amo, amare, amavi, amatum. Ferre is the most irregular Latin verb I know).

Here are examples of words that survive, derived from this irregular verb:

• Fero -- apart from scholarly words like circonferenza, there's Lucífero (="Light-bearer")
• ferre -- Iron was called in Latin (and thence Italian) ferro because it was weight-bearing or just heavy
• tuli - can't find any; as the most irregular form of the most irregular verb, I doubt if it has any derivatives
• latum -- lato, meaning broad/wide/extensive, and words derived from that, eg latifondo

(I‘m not sure about   this   last one. This, from Etymonline, sv flat suggests  another possible derivation for lato:

flat (adj.)c. 1300, "stretched out (on a surface), prostrate, lying the whole length on the ground;" mid-14c., "level, all in one plane; even, smooth;" of a roof, "low-pitched," from Old Norse flatr "flat," from Proto-Germanic *flata- (source also of Old Saxon flat "flat, shallow," Old High German flaz "flat, level," Old High German flezzi "floor"), from PIE root *plat- "to spread."

And this Italian source suggests yet another. I should have paid more attention in my History of Italian lectures.)

The way con|m- words works reminds me of the way im-|n- words work -

• impossible, immoral, imbibe (combat, commemorative, compact)
  but
• infant, invective (conference, convection)
  
  but
  
• illegible, irreverent (collect, correct).

But whereas in- is what linguists call "a productive affix" (one used by current speakers, throwing up phonological changes on-the-fly as they form  negatives – and, incidentally, with attendant  problems for speakers of ESOL – particularly speakers of languages with different rules, such as the Spanish inmoral), the changes thrown up by Latin CUM- are lost in the etymological mists –  of current interest only for people who want to get the distribution of ms right in commemorative. (In fact, when a newer source [not Latin, as is the case with, say, condescension]  is involved in adding the prefix, these rules don't apply: co-dependant. not *condependant)

There are things to  be doing though.

b