Answers to the Blog Search Challenge
Last Friday, I posted a series of clues to words found in the King James Bible and challenged you to find them using Accordance wildcard symbols and search commands. We promised to give anyone who successfully completed this challenge a $30 credit toward their next Accordance purchase. About fifty people responded to the challenge, so we gave away nearly $1500 worth of credit! Even better, most of the participants reported that the exercise had taught them something new.
For the rest of this post, I want to give you the answers to each question and discuss the search methods used to find them.
1. A five-letter palindrome in Ezra
A palindrome is a word which is spelled the same whether it is read backwards or forwards, like the name Hannah. To find a five-letter palindrome, you would use five character wildcards (i.e.: the question mark) and then indicate that the fourth letter needs to be the same as the second letter, and that the last letter needs to be the same as the first letter. Like this: ????(=2)?(=1).
Placing parentheses after a character wildcard enables you to specify which characters that wildcard can represent. By placing an equals sign followed by a number in the parentheses, you can specify that this character must be the same as the character represented by the number. Note that the number must refer to another character wildcard. Thus, =2 refers to the second question mark in the search.
To constrain this search to the book of Ezra, you could set up a permanent range in the range pop-up of the Search window or use the RANGE command to define a temporary range, like this: ????(=2)?(=1) <AND> [RANGE Ezra]. The word found by this search is Aziza.
2. A place name with 3 sets of doubled letters (like MiSSiSSiPPi)
This search could be accomplished with the same technique as the last search, but it required the use of some asterisk wildcards to account for characters which may or may not appear before, between, and after the doubled letters. Searching for *??(=1)*??(=3)*??(=5)* finds the place name Kibroth-hattaavah (note the doubled h, t, and a). This search also finds regular words like "rottenness," "lookingglasses," and "barrenness," which is why I asked you to give me the place name.
3. The longest word in Isaiah (18 letters).
Finding this word required using my clue that it was 18 letters long. Eighteen letters can be found using 18 question marks. ?????????????????? <AND> [RANGE Isaiah] finds Maher-Shalal-Hash-Baz—a name I've suggested for each of my five children, but to which my wife has never agreed. (I just can't figure it out!)
4. A name with 4 consecutive vowels
This search required that you constrain the question marks to a certain set of characters. Most people used *?(aeiou)?(aeiou)?(aeiou)?(aeiou)*, while a few also included a "y" in the parentheses. Either approach would return the name Reaiah (or Reaia).
5. A name with one letter repeated five times
This search seems to have presented the greatest difficulty for most people, and it actually exposed an issue with the way we handle wildcards. The search which I was looking for, and which most people came up with, is *?*?(=1)*?(=1)*?(=1)*?(=1)*. This search returns a single word, Aramnaharaim, which has the letter "a" repeated five times.
Then someone pointed out that the word returned by a previous search, "Maher-shalal-hash-baz" also has five a's, and I realized that the search I had come up with was missing some words it should have been finding!
The problem is that the initial asterisk, which should account for any characters before the repeated letter, is basically being ignored. So "Aramnaharaim," in which the first letter is repeated, is found, but words in which a second or third letter are repeated, such as "Maher-shalal-hash-baz," "Zaphnathpaaneah," or "possessions" are being missed. Obviously, we'll get this fixed as soon as we can. Fortunately, our error didn't keep anyone from being able to find an answer to this question.
6. A six-letter word in which the first three letters are repeated (that is, in the pattern abcabc)
Most people found this one easily. Searching for ????(=1)?(=2)?(=3) returns the word murmur.
One intrepid soul went even further than this. He realized that the search above would not necessarily find the pattern abcabc, since it might also find patterns like aacaac, abbabb, aaaaaa, etc. In order to find only the pattern abcabc, you would have to make sure that the first three letters did not match each other. I didn't have that level of specificity in mind when I phrased the question, but it does make for an interesting challenge.
The user who tried this used the search ??(-1)?(-1,2)?(=1)?(=2)?(=3), and because it found "murmur," he assumed it had worked. But this search didn't exactly do what he intended. First, because he didn't use the equals sign in the parentheses after the first two question marks, he actually specified that the second character could not be the number "1" and the third character could not be "1" or "2". To specify that one character cannot be the same as another, you have to use the equals sign followed by a minus. Thus, ??(=-1) will find any two-letter word in which the second letter is different than the first.
I could find no way to specify that a given character could not be the same as more than one other character. In other words, ??(=-1)?(=-1=-2) or something similar will not specify that the third character must be different than both the first and second. Accordance will not return an error for that search, but it just seems to ignore everything after the first equals expression. In other words, ??(=-1)?(=-1=-2) is no different than ??(=-1)?(=-1).
The only way I found to accomplish the more specific search this user attempted was this: ?(=-2)?(=-3)?(=-1)?(=1)?(=2)?(=3). By saying that each of the first three characters could not equal one of the other three, and making sure that each excluded a different character, I was able to find a true abc pattern. Then I just needed to specify that the pattern was repeated for the last three characters. The result, of course, was the same, with "murmur" being the only hit.
In the end, this intrepid user made a valiant attempt at a very sophisticated search, and challenged me to push symbol searching in Accordance even farther than I realized it could be taken!
7. The most common word used at the beginning of a book
8. The most common name used at the beginning of a book
These two searches, along with the final two, were designed to get you using the FIELD command. The FIELD command lets you specify the proximity of a word to either the beginning or the end of the current search field. The "field" is what we call the literary unit you choose to search using the Search within every pop-up in the More options section of the Search window. To find the words at the beginning of a book, you would select Book from the Search within every pop-up, then enter the search * <WITHIN 1 WORDS> [FIELD Begin]. By clicking the Details button and checking either the Analysis, the Analysis Bar Chart, or the Analysis Pie Chart, you could then see that "The" is the most common word used at the beginning of a Biblical book, and "Paul" is the most common name.
By the way, this challenge exposed another Accordance bug we need to fix. In Accordance 8.1, we decided to place the Sort pop-up menu directly on the Analysis window, rather than forcing you to open the Set Analysis Display dialog to change the sort. Unfortunately, we forgot one thing. You can only choose the Count Down and Count Up options when Key numbers are not displayed. In the dialog, if you try to sort in one of these ways with Key numbers displayed, Accordance gives you an error message. But we forgot to give that error message when you choose those sort options from the pop-up menu on the Analysis window itself. If you tried to change the sort for the KJVS and were surprised when Accordance did nothing, I'm sorry for the confusion, and I'm told it's already been fixed for the next update.
9. The most common word used at the beginning of a sentence
For this question, you could keep the same search and just change the Search within every pop-up from Book to Sentence. Again, checking the Details showed that the most common word used at the beginning of a sentence in the King James Bible is "And"—a fact which must drive English grammarians crazy!
At this point, let me mention that a few users took a different approach to finding the first word of a book or sentence. Rather than using the FIELD command, they opened a Construct window and used the PLACE item. The PLACE item lets you specify the place a given word occupies in the field. A place of "1" will find the first word in the field, a place of "10" will find the tenth word, and so forth. When searching for the first word, this method is just as effective as using the FIELD command. However, these users had to come up with another method to answer the final challenge.
10. The most common word used at the end of a sentence
The final challenge was to find the most commonly used word at the end of a sentence. So with the field still set to Sentence, it was just a matter of changing the search to * <WITHIN 1 WORDS> [FIELD End]. A check of the Details showed the word "him" to be the one most used at the end of a sentence.
Other users met this challenge by searching for a word followed by a sentence-ending form of punctuation, such as a period, question mark, or exclamation point. Either method was acceptable.
All in all, this searching challenge was a huge success. Everyone, it seems, learned something about searching in Accordance . . . including me! We gave away a lot of credit which I hope you all will put to good use. And we even found a couple of bugs that needed fixing. Best of all, everyone seems to have had fun with it, which was the very thing I was after.