Euphonic rules for naming

by Rick Jelliffe

The issue of the best rules for naming came up recently. I think the way of the future for real human readability is euphony. We want beautiful Code, so why not beautiful markup? With the current discussions on XML 1.0 version 5 and its naming rules buzzing around, euphony has surely never been a more critical issue.

So what might rules be? "Jelliffe" is an old spelling for jolly. My parent's traced the familly tree back, they think, to Bosham, England a thousand years ago, which is when King Canute was there.

One of the interesting parts of the tree is how often the same names come up: John Jelliffe, William Jelliffe and so on. Common enough names, but relentless and I have often wondered what the euphonic principles were. I have been looking at various a genealogical sites recently, and I think I have cracked the rules.

We start of by discarding wives names. I am not sure many people picked their spouse because of their first name, as in The Importance of being Earnest (though there is an Earnest.) Here are the rules as I see them:

1) Any name that starts with J sound. But this must be a single syllable name (or single plus a diminuative.) John. Jane. James/Jimmy, Charles, Jessie

2) Any name that has an internal rhyme or sound in the same position. Emma, Thelma, Elton, Benton, Benson, Benjamin, Frederick, Fred, Terence, Preston, Elsie, William, Mildred, Helen, Ella, Illah (my favourite), Lillian, Ralph?, Pauline, Wellford, Edyth, Edwin, Jesse, Ethelyne, Elinor, Elward, Patrick

3) Any name with the same rhythm making dash dot dash dot: Catherine, Martha, Robert, Clara, Hazel, Joseph, Edith, David, Taylor, Martin, Emma, Roger, Myrtle, Walter, Waltby, Ella, Vincent, Arthur, Sarah, Eva, Richmond, Zada, Jesse, Elward, Rufus, Walter, William, Zalmon, Mahala, Thirza, Stebbins, Fannie

4) Any name with a soft sound that matches the ffs and soft J: Catherine, Martha, Howard, Edith, Nathaniel, Smith, Fern, Phoebe, Arthur, Eva, Virginia, Willis, Owen, Fond, William, Fannie

Of the rest, there is mob that seems to have surnames as their first name, in the American fashion (Burr). But these rules catch most of them. Some odd runs in New Jersey: Gustiss, Freylinghysen.

So a name like Estha really fits the euphonic bill: 2, 3 and 4.

And I guess it makes sense of the first name Luzon which crops up: rules 3 and 4.

So for people who are passionate about beautiful documents, I suggest that attribute names and namespace prefixes should be picked to be euphonic with the element names. We needn't, indeed we mustn't and shan't, put up with names picked because of some arbitrary theory of utility or historical accident, someone's mere whim or assertion. We need the science of euphony to get really beautiful markup and proper human memorability (surely a better goal than just readability?)


2008-02-20 06:33:50
So now the XML conference keynotes will be poetry readings instead of open source rants?

Well that would be a change.

Geoffrey Spear
2008-02-20 08:10:15
"XML 1.0 version 5" doesn't exactly inspire confidence that anyone involved with the project has a clue about naming things.
Bob Jolliffe
2008-02-20 13:34:06
Hi Rick

Glad to see you taking time out to muse on the importance of names and naming. My wife's surname is Byrne and we anguished for some time over the kid's being "Byrne-Jolliffe". We decided against for obvious reasons.

But names are important and socially connect us to people and things.

Any thougts about the name "Office Open XML"? There's been a lot of observation that its confusing. Including from South Africa. ECMA's reponded: No its not. Not their most responsive of responses.

Kind of like Jellife and Jolliffe maybe.

If you were to pick a new name what would you suggest?


Rick Jelliffe
2008-02-20 18:21:13
Bob: Is that the South African Bob?

On the name Office Open XML, there is the name of the technology, which is a branding issue outside the scope of ISO and cannot be helped, and then there is the name of the standard, which should be generic, descriptive and not have product names. "Office" is difficult, because AFAIK it is not actually an MS trademark and you would expect some leeway for fast-tracked/PAS standards to keep their identities.

However, I don't see that there is any downside to the standard having a more generic and descriptive name. In fact, this is one of the issues I raised, which got incorporated into Standards Australia's comments. But it is not super high on the list of priorities, I think: I tend to rank issues by importance as 1) conformance and schemas and references, 2) normative technical substance 3) normative editorial, 4) normative material like titles, intros and scope statements, and last of all non-normative
material and examples.

2008-02-21 23:56:57
Geoffrey Spear:
""XML 1.0 version 5" doesn't exactly inspire confidence that anyone involved with the project has a clue about naming things."

Ecma 376 gave the impression that no one at MS considered naming worth much thought. Much to the peril of anyone who wants to implement DIS 29500. Personally, I consider it the task of the standard's writer to ease the burden on those who have to implement it. And there are a lot of tag names in DIS 29500 to understand and very little in semantics in the description.

Eg, I can't see the system behind:
outerShdw, ActiveWritingStyle, attachedSchema, documentType, docVars, endnotePr, hdrShapeDefaults
(many children of the same parent tag)

The problem is then for new implementers to decode all these "mnemonics" which look to me like C++ variable names.

Rick Jelliffe, Geneva
2008-02-24 08:12:39
Phil: There have been exactly ZERO "proven cases of corruption".

Shame on you.

This kind of glib and ruthless smearing of people and institutions, where any and every time someone gets surprised or disappointed they claim it is prima facie evidence of corruption, despite not have any actual evidence of it, goes beyond panic, Chinese whispers, sour grapes or monumental self-righteousness, but starts to look mischievous, libelous and pathological. Fortunately, it is only a tiny, touchy minority, but you have no idea how much unnecessary disrepute it brings to your cause.

We have to expect SNAFUs on a project of this scale, from all parties: you have lots of people with very little experience, which makes them both prone to making mistakes on the one hand and inflating the importance of the mistakes on the other.

Any future comments relating to DIS 29500 this week will be removed, and please don't be wankers and try to draw me into such discussion. The purpose of this item, apart from its fascinating subject matter and deathless prose, was to point out that we can get very dogmatic about markup issues that are really matters of aesthetics and local or even personal idiom. (Geneva is a quite good city for contemplating the ins and outs of dogmatism!)