Matching Socks

I had a dream a few nights ago in which a young woman of my close acquaintance tilted her head to one side and looked at me quizzically and asked why I had to match all the socks. It came to me in my dream that I do not need to match socks. When I woke up I decided that I would continue to match socks just the way I always do, with partial success.

However, when it comes to codepoints and search strings you really have to match socks, or find away around it.

Here is the word stenochoreo (2 Cor. 8:4a) in three different sets of codepoints.

1. στενοχωρέω combining accents

2. στενοχωρέω precombined accents polytonic

3. στενοχωρέω precombined accents monotonic

The first is from e-sword and I took it from Scripture Zealots blogpost, the second is from zhubert, and the third is from the LSJ lexicon in the new Perseus Project at PhiloLogic.

I always say used to say that they just “look different.” But now I am over fifty and they all look the same. However, they all have different codepoints for the epsilon + accent. They will not work across platforms. They are mismatched socks but they do a good impersonation of being a match. I wish my socks were half this clever.

Now here is where I always say to my kids: do you want the long story or the short story? They always pick the short story. So, that’s what I’ll do for now. But if you like I can post the long story another day.

Here is my advice for searching. Always use the simplest choice possible. Do not mess with the accents if you can help it.

1. To access the LSJ lexicon at the original Perseus Project>classics>other tools & lexica>dictionary entry lookup. Choose Greek from the menu and follow the specific instructions from the keyboard display. For στενοχωρεω, you will type “stenocho^reo^”. The accents are used to create the long vowels. You do not have to indicate Greek accents. Now type stenocho^reo^ into the dialog box on this page and submit query. This should be the result. Its slow. I always try it a few times, back and forth. Match a few more socks, etc.

You can reconfigure these pages to look like Greek rather than Latin but I do not spend the time to do that.

2. To access the LSJ lexicon at Philologic>Reference Works>Search Liddell and Scott’s Greek English Lexicon Now here is the trick. DO NOT mess with the accents.

(From what I can see this lexicon uses a mix of accents – tonos – and not oxia, along with varia and circumflex. If anyone would care to explain to me what kind of system that is and if there is a keyboard which accommodates such a thing, I would be interested. I really don’t understand it.)

Go to the Greek inputter, or use your favourite Greek keyboard – I use the bundled with MS version, and type in the word stenochoreo with the Greek inputter. As I said, NO accents. στενοχωρεω. You can just type this in the box by choosing the English letter that looks most like the Greek letter. For στενοχωρεω I typed stenoxwrew.

Now put στενοχωρεω into the LSJ at Philologic and choose the “No accents or breathings” option. The page should look like this. Choose στενοχωρεω from the list and the results are here.

You will notice that the LSJ entry at the original Perseus Project has hyperlinks and the one at Philologic does not. I typically still use the original Perseus Project, in case I want to use the hyperlinks.

3. For searching Zhubert, you need full polytonic Greek. Now you have to mess with the accents. Go to the Greek inputter and type in the word using all the accents. You will need to click on the “Greek letters” tab and carefully select the letters with the correct accents. στενοχωρέω

Now cut and paste, or try to reproduce this, and paste it into the wordfinder on Zhubert. Click “find”. You should see a page with 6 results. None of them are for 2 Cor. 4:8. At this point technology breaks down. Zhubert has parsed the verb in 2 Cor. 4:8 as a middle voice verb but this is not a headword in the LSJ. It will not compute.

There are simply times when there is no way around not having a strong foundation in Greek grammar. I do not think that software searches can ever make up for studying a sizable amount of Greek. This should enable you to find most words in the LSJ, but not all.

Anyway, now you know why my kids always choose the short story. Its still long. Please ask me to explain whatever is not already clear as mud.

13 thoughts on “Matching Socks

  1. Peter Kirk says:

    If those three look the same to you (they don’t to me, but only if I take my glasses off and peer at them) that is as it should be. According to Unicode standards the three forms are canonically equivalent and so should be rendered identically, and any one of them should match any other in any search. If this doesn’t happen it is because we are using obsolete technology like Windows XP. I would be interested to know if they actually appear any different with Windows Vista.

    Sorry to bore those of you who come here to learn about the Bible and not to sit in on a discussion between Unicode geeks!

  2. mgvh says:

    Instead of using LSJ off the web (which can be quite slow at times), you really want the free Diogenes program which puts the full LSJ (and the Lewis-Short Latin) on your computer. Info here:
    You can choose whether to enter searches using Unicode, Perseus code, or Beta code.
    As for the forms of epsilon + acute, as you note, the first is combined > it's actually two characters, an epsilon and an acute combined. (03B5 + 0301)For the second, it's using epsilon with oxia. (1F73)For the third it's using a precombined epsilon with tonos. (03AD) As Peter notes, they should all be treated the same, but for now, they certainly are not in MSWord.

  3. Bob MacDonald says:

    Hah – I am glad I am learning Hebrew. Sometimes search technology makes you think you can learn everything at once. The long story is better. Good things in the story seem to take time.

  4. J. K. Gayle says:

    Your good point, Bob, makes me remember something Ken Pike wrote. Here’s his last paragraph of the essay “Why Poetry?”

    “Except we become as little children, we can neither learn a new language without a bad accent, nor become charter citizens of heaven–nor experience that multiple fullness of the n-dimensional experience, which poetry tries to help us capture. Poetry compacts life in language–as an oak forces all its eternal blueprints into one nutshell.”

    Suzanne, Pike could have just as easily said, “as a story told short is the blueprint for long telling.” And he usually ended his commentary on his monolingual demonstrations with a poem. I do remember that Dr. Seuss book on Socks (it was a rhymer “Socks on Knox and Knox in box…Who sews whose socks? Sue sews Sue’s socks. Who sees who sew whose new socks, sir? You see Sue sew Sue’s new socks, sir.”) which helped me read more here from you about codepoints and search strings too. Sorry for the silliness. Thanks for the post.

  5. Scripture Zealot says:

    That was a good explanation. Thank you.

    I will try the stand alone version also.

    I tried three different Greek fonts in e-Sword and they all come up the same so that’s probably not what’s causing the accents. It would be nice to just copy and paste from there.

    I’m at the beginning of the short story which will take a long time and will see if I get to the long story. The thought of Hebrew makes me sweat.

  6. Suzanne McCarthy says:

    Jeff (SZ)

    That’s the problem. You cannot just copy and paste. Although Peter indicates that they are “canonically equivalent” the text from e-sword did not get results in PhiloLogic. So, all this technology is in some sense obsolete. That’s my take. It is not a matter of fonts, but of actual codepoints, which need to be analysed to be seen (long story).

    Mark, (mgvh)

    I value the hypertext at the original Perseus. I am aware of the different codepoints but I do not understand why PhiloLogic uses tonos and varia, instead of oxia and varia. That is, is there such a thing as a keyboard which produces the text at PhiloLogic? I do know that there were converters created for Unicode Greek that produced mixed text such as this, but I forget the details.


    Hebrew may have a few issues, but overall I think it is much less complicated to compute in Hebrew.

  7. Peter Kirk says:

    Suzanne, as I understand the Unicode guidelines the “canonical” and therefore recommended code points use tonos rather than oxia. That is, the standard conversion to normalised form D (decomposed) converts oxia to tonos, whereas normalised form C (composed) uses the precomposed characters where possible. So PhiloLogic is doing the right thing, whether they realise it or not!

    I note that in many fonts including the one with which I view BBB tonos is incorrectly rendered with a vertical line, which is apparently never considered correct even in monotonic Greek. Tonos should slope to the right, like oxia, but more steeply than the acute accent in French etc. But that is a matter for font manufacturers.

  8. Suzanne McCarthy says:


    I am happy to know that PhiloLogic is doing the right thing, but I don’t believe that there is any codepoint for “tonos and psili” so at that point there would have to be “oxia and psili”, in which case, the text uses both tonos and oxia. Does that make sense?

    For example, ἄρχω in PhiloLogic uses U+1F04 : GREEK SMALL LETTER ALPHA WITH PSILI AND OXIA.

    To tell the truth, I have always keyed in U+1F71 : GREEK SMALL LETTER ALPHA WITH OXIA and not U+03AC : GREEK SMALL LETTER ALPHA WITH TONOS and the same for the other vowels – on the MS Greek polytonic keyboard. What does your keyboard produce, or is there a choice?

    Well, that is why I don’t like the long story. Whatever, all I know is that Zhubert, PhiloLogic and e-sword do not match and one cannot cut and paste text from one to another.

  9. Peter Kirk says:

    You are right, from memory, there is a lot of inconsistency here. There is no “tonos and psili” because psili is used only in polytonic Greek which has oxia rather than tonos – except that oxia and tonos are really the same thing.

    You should certainly use U+03AC rather than U+1F71. The latter is anyway likely to be automatically converted to the former.

    I don’t have a Greek keyboard, only a Hebrew one which is much simpler! To key Greek I use BabelMap, or copy from a Bible text.

  10. Suzanne McCarthy says:

    But oxia and tonos have different codepoints. This all reminds me of how I fell in love with Unicode in the first place – for its imperfections.

    You should certainly use U+03AC rather than U+1F71. The latter is anyway likely to be automatically converted to the former.

    I don’t think it is. My Microsoft bundled Greek polytonic keyboard matches Zhubert but not PhiloLogic. There is no automatic conversion that makes it work. My concern is that there might be people who think that these searches just don’t work “for them,” when it is the codepoints that don’t match.

    This is why the Greek inputter is so great. You can type along and then when you need an accented letter you can choose it from the chart. Much faster than Babelmap, although that is essential for analysis.

  11. David Ker says:

    Unicode…imperfections! Wow, you guys are deep.

    Thanks for dressing up an otherwise boring post in funny socks.

    I do everything by cut and paste for the proprietary software I was given when I became a BT. But my aptitude for Greek isn’t even in the same league with yours.

    Peter, I’m amazed at your geekery regarding Unicode.

  12. David Ker says:

    I actually know some other people on that list. They’re normal like you in a weird sort of way.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s