Using the Chicago Homer
The Chicago Homer is a bilingual database of the Iliad and Odyssey as well as the poems of Hesiod and the Homeric Hymns. Except for fragments, the Chicago Homer includes the entire extant corpus of Early Greek epic. And as an electronic medium, this site not only supports the types of searches you would expect from such an environment, but because it stores much of its information in a relational database, it also supports many searches not ordinarily associated with a text archive.
"Using the Chicago Homer" is one of three help documents and gives you a hands-on tour of this site. We strongly recommend that you take the time to read the longer and more analytical introduction Understanding the Chicago Homer. You will take better advantage of this site's query potential if you have a firm sense of how the data are organized and accessed. Finally, there is a tutorial, What can you do with the Chicago Homer, which develops in some detail two scenarios of use, the first for an intermediate student of Greek and the second for a Greekless reader.
Two Special Notes
1. Basic design choices and their trade-offs
The Chicago Homer loads a stack of "pages" into your browser that are layered invisibly on top of each other (i-frames). You use special tab controls to move between them. This has the advantage that you can move from any page to any other page directly without going through the intermediate pages. The flip side of this advantage is that it conflicts with ordinary habits of web browsing, where you use the back button to move forward and backward. The search and browse pages open in a special window that lacks generic browser controls. If you want to adjust browser controls (e.g. font size), you must return to the Home window.
Think "up and down" rather than "forward and backward" as you navigate The Chicago Homer.
The layered design of The Chicago Homer works best with Internet Explorer under Windows NT, 2000, or XP, and with the Safari Browser under Macintosh OSX. With earlier combinations of browsers and operating systems you are likely to experience glitches of various kinds. These are detailed in the Technical Requirements section. All of them stem from problems the browser has in clearing or refreshing a particular layer, and many of them can be worked around if you manually refresh the screen by moving the scroll bar up and down.
1. Displaying Greek fonts
The Chicago Homer displays Greek in Unicode only. Specifically, it follows the recommendation of the World Wide Web Consortium (W3C) and uses Normalization Form C, in which polytonic Greek is represented by means of precomposed characters. Support for this display comes "out of the box" for Windows 2000, XP, and Macintosh OSX if you choose a Unicode font that includes the extended Greek character set, , such as Palatino Linotype, Arial Unicode MS or Apple's Helvetica. Your browser needs to be set to UTF-8 encoding. It may or may not automatically adjust itself.
If you have Windows 98, you may not have not have a font that includes the extended Greek character set. You can download David Perry's free and very fine Cardo font and put it in your fonts directory, typically C:\Windows\Fonts.
On the Macintosh, Unicode support is spotty for OS 8.5 and 9. David Perry has a version of Cardo for Mac OS 8/9 but he does not guarantee it will work fully. Your best bet on earlier versions of the Mac is probably to stick to the transliterated display. As I argue more fully in Understanding The Chicago Homer, this is not necessarily a second best and is in some respects closer to the orthographical conventions that Plato or Sophocles were familiar with.
The Basic Control Tabs
When you open the Chicago Homer, you see at the top of the screen a fringed bar with the tab settings
SEARCH† BROWSE† HELP† CORRECTIONS† OPTIONS† HOME.
You control the program with this navigation bar, which is always visible. Clicking on any of the tabs will bring up a particular page. The tab will turn black, which marks the active page. Note also that the SEARCH and BROWSE tabs bring up different sets of secondary controls.
BROWSE is the gateway to everything that begins with reading a section of the text. SEARCH takes you to operations that begin with looking up terms or looking for phenomena that meet specified constraints.
Click on BROWSE and watch its color change to black, indicating that the browse mode has been activated. You now see a line that lists the available texts. Clicking on any of them brings up a third control line that lets you retrieve a particular section of the text by book and line number.If you specify nothing, the beginning of the text is automatically selected. You must click on RETRIEVE to activate your choice. The PREVIOUS and NEXT buttons on the right of the third control line let you move forward and backward within a particular book or poem. Note, however, that you cannot use this method to go from the end of one book to the beginning of the next.
Whenever text appears in Browse mode, every Greek word carries an invisible link and clicking on it will produce a grammar table and a frequency table in the right margin.
The frequency table gives information about the "lemma," or dictionary entry, of the word form, telling you how often all forms of that lemma occur in each work and in the corpus of Early Greek epic. For the Iliad, the Odyssey, and the four major Homeric Hymns, it also provides information about the relative frequencies in narrative and speech. Relative frequencies are expressed as occurrences per 10,000 lines. The decimals in these frequencies do not add useful precision but they do mark the derived status of the figures.
The grammar table lists the word form, the lemma, and a description by word type and the appropriate categories of tense, mood, voice, case, gender, person, and number.
Several elements in the grammar table contain hyperlinks that let you move in three different directions:
Whether you click on the word form or the lemma, you notice a change of display. SEARCH turns black on the top row of the navigation bar, and in the bottom row you see that CONCORDANCE has turned black.
The text has disappeared, and instead you see the concordance output of the search that you triggered by clicking on the word form or lemma. This output consists of a single line for each hit. In this display, the words are not linked to the grammar feature.
If you want to see the line in context, click on the line number to the left of the text. The site now returns to browse mode and takes you to the text. You can return to the concordance output by clicking on SEARCH. You can then look up the full text of another concordance line, and in general you can shuttle easily and quickly between word, keyword in context, and full text. Each "layer" of search mode and browse mode, regardless of whether it is on top or not, will keep its content until you change it.
The Options Tab
The OPTIONS tab opens a control window that lets you specify values for the display of text and for some other operations of the program. The default settings are as follows:
If you change the settings, you must return to BROWSE and click on the RETRIEVE button for the new settings to take effect.
The text display settings are self-explanatory. The other settings allow you to perform different kinds of searches and are discussed separately in the next sections.
English-Greek Translation Linking
When "English-Greek translation linking" is on, some words or phrases in the Lattimore translations of the Iliad and Odyssey appear as links. Clicking on any of them will produce a list of one or more Greek words in the right margin. This list consists of all the Greek nouns or adjectives that are translated by the English word or phrase.
Try this for the English word "anger" in the first line of the Iliad. Ten words appear in the margin in their lemmatized, or dictionary entry, form. If you don't know Greek, you can still match spellings. Thus you will see that the Greek word translated by "anger" in this line must be "mÍnis," because the form "mÍnin" appears in the corresponding Greek text.
The lemma "mÍnis" is itself a link. If you click on it, the search layer moves to the top and the SEARCH tab turns black again. In the second row of the control panel CONCORDANCE lights up, and in the main window you see a concordance, or "keyword in context," display of all the lines in which the lemma "mÍnis" occurs. Clicking on the line number to the left of the line will take you to the full text of the passage in browse mode. Clicking on SEARCH will return you to the list of occurrences of "mÍnis." Whether or not you know Greek, you can easily read the opening word of the Iliad in the context of all its occurrences.
The acronym "LSJ" appears as a link before each Greek word in the margin. This refers to Liddell-Scott-Jones, the most authoritative dictionary of Greek, and clicking on the link will take you to the entry for the word in the electronic version of the dictionary on the Perseus Web site.
English-Greek translation linking is a pretty rough tool with substantial limitations, which are explained in the relevant section of "Understanding the Chicago Homer." But for many purposes it will allow a reader with little or no Greek to search or navigate the original text through the translation.
Repeated Phrase Markers
The Principle Behind the Feature
The most distinctive feature of the Chicago Homer is its ability to provide information about repetitions. Not just the famous phrases like "rosy-fingered Dawn" or "swift-footed Achilles," but any and all instances where a given sequence of two or more words is repeated elsewhere in the corpus.
Imagine a listener in antiquity at a performance of a Homeric poem. Such a listener would be quite familiar with the conventions of the poetry, and particular phrases in a given stretch of lines would resonate with the contexts of their occurrences elsewhere. It is not easy for the modern reader of Homer to hear those echoes, and translations are not very reliable in reproducing them. The repetitions feature of the Chicago Homer allows you to simulate the competence of the ancient listener: the text with its repetition markers points to places where echoes are possible, and following the links of repeated passages is a way of navigating the neural networks of bardic memory.
How It Works
When you enable "display repeated phrase markers" in Options, you are given some additional choices. You can:
The first three items have default settings. Ignore them for now, but specify the frequency as greater than 4. Select Iliad, book 1, line 1 on the control panel above and click on the RETRIEVE button.
When the text appears you will see that some word strings are enclosed by square brackets with numbers. These word strings are the repeated phrases that meet the search parameters you just specified under the "display repeated phrase markers" option. The numbers keep a running count of the repeated phrases on your page.
In the first line of the Iliad there is the word string "PÍlÍÔadeŰ AchilÍos." Even without knowing Greek you can guess that this is probably the Greek for "Peleus' son Achilleus." If you click on the bracket enclosing the phrase, the application changes from browse mode to search mode, and REPETITIONS turns black on the bottom row of the navigation bar.
Below the navigation bar you now see the form on which the Chicago Homer reports the results, or "hits," of repetition searches. In this case you will see only one phrase variant, "PÍlÍÔadeŰ AchilÍos." Clicking on it will take you to a list of citations for that phrase, and clicking on any citation will take you to browse mode and the full text in which that instance of the repeated phrase occurs. Clicking on SEARCH will return you to the list of repetitions. You can shuttle between repetitions and full-text display in the same way in which you shuttle between concordance and full-text display for grammar-linking or the English-Greek index.
Try the same thing for the phrase "dios Achilleus" in Iliad 1.7, which you might guess is the Greek translated by "brilliant Achilleus." The repetitions form now displays the following:
dios Achilleus (55)
List of all occurrences (57)
Clicking on the link for the list of all occurrences will bring up a list of citations for all occurrences of all phrase variants. Clicking on the link for a specific phrase variant will take you only to the citations for that variant. In this case, you notice that one variant heavily dominates and the other is quite rare.
Filtering Repetitions in Different Ways
In the examples so far you have seen the text with a setting that filtered out repetitions unless they occurred at least five times. Return to OPTIONS and under "display repeated phrase markers" remove the number 4 from the frequency box. Return to the text by clicking on the RETRIEVE button. You see immediately that there are many more repetitions and that some lines are so full of square brackets that it is difficult to make sense of what is going on. If you want to analyze a particular verse paragraph for its repetitions, it will probably work better if you do it in several passes, moving from rare to common repetitions or from short to long ones.
The SEARCH tab on the navigation bar is the gateway to all operations in which you look up a word or look for words and phrases that meet constraints of your choosing.
Once you click on it, SEARCH turns black, and the buttons on the second row of the control panel change to the following:
QUERY FORM ††WORD LIST†† CONCORDANCE†† REPETITIONS
Clicking on QUERY FORM takes you to the form on which you specify search constraints. You notice that the top frame now wraps around on the left side of the screen, which becomes a search sidebar. This sidebar contains instructions and navigational aids for specifying search criteria in the query form in the main part of the screen.
You execute a search by clicking on the Search button at the bottom of the sidebar. Clicking the Reset button to the left of the Search button will clear the contents of the query form and permit you to formulate another search from scratch. If you do not click the Reset button, new selection criteria will be added to the criteria of an earlier search. This is a feature rather than a bug, but it is also the source of error and can lead to unintended results.
Whatever the nature or complexity of your search, it will always produce a report in two parts. When you click on the Search button, the query form disappears, and depending on the nature of your search, you will see either the word list or the repetitions form, with WORD LIST or REPETITIONS turning black on the navigation bar. While the search itself is running, the previous search report will still appear in the main window. When the search is done, the main window will display the three parts of the new search report.
Word forms, lemmata, and repeated phrases in the search reports are followed by numbers in parentheses. These numbers always refer to counts in the total corpus, so they are not generated "on the fly" by the constraints of a particular search. On the other hand, particular search constraints will govern the lines that appear in concordance output.
Specifying a Search
Searches in the Chicago Homer are not run against the text of the epics, but against a set of tables derived from them. In addition to looking up particular words, you can look for unknown words or phrases that meet certain constraints, such as words spoken by Penelope and occurring fewer than five times.
A search report will give you the results of a search based on the following constraints separately or in combination:
Of the various constraints, only the choice between words or phrases is mandatory. You make this choice at the top of the search side by clicking on the WORDS or PHRASES tab. You will notice that the query form changes with your selection.
If you specify a search for words, you have the option of seeing the search results at the lemma level or at the more granular word form level. You make your choice at the bottom of the search sidebar, where it says:
Report results at the level of:
The default setting is LEMMA.
As for particular search constraints, you must specify at least one, but you may specify or ignore any particular constraints separately or in combination.
The query form is quite long and complex. You can scroll up and down, but you can also use the headings on the search sidebar to go directly to relevant sections of the query form.
Particular search constraints are discussed in the next section.
The Search Term box at the top of the query form lets you enter one or more search terms. If you don't want to constrain your search with particular words, don't enter anything.
While the Greek text of the Chicago Homer is displayed in Unicode, Greek words and phrases in the tables are stored in beta code and in transliterated form, allowing you to search for either. You can enter terms in transliteration or in accented or unaccented beta code. You can also specify a particular word form or enter the lemmatized (dictionary entry) form, which will retrieve all inflected variants of the search term. The default settings are transliteration and lemma.
Enter multiple terms with blank spaces and no punctuation between them.
You must use the following "workarounds" for entering diacritical marks used in transliteration (your browser can display diacritical marks but does not recognize them as an input):
o^mos for Űmos
The Chicago Homer supports wildcard searches. The most commonly used wildcard character is the percentage sign, which stands for any combination of characters, including none. Thus "an%" retrieves any character string beginning with "an"; "%an" retrieves any character string ending in "an"; and "%an%" retrieves any character string that contains "an," including "an" itself.†The underscore sign (_) is used as a wildcard character for a single character. This can lead to complications with transliterated search terms. Thus "_mos" will retrieve "emos," but it will not retrieve "Ímos" or "Űmos," because the first letters of these words are stored in the data tables as the entity references "ê" and "ô".
If you are looking for repeated phrases, the search term box will include an any/all choice, corresponding to Boolean OR and AND. Keep in mind that Boolean AND is a more restrictive choice than Boolean OR and will always retrieve a smaller result set. If you enter the Greek words for "man" and "woman" in the search box and choose all (the default), the search will be limited to phrases that contain both the words "anÍr" and "gunÍ." If you choose any, the search will return all the phrases that contain "anÍr" and all the phrases that contain "gunÍ."
Frequency and Length
Frequency constraints operate almost identically with words and repeated phrases. The default setting is "one or more" and includes some other variables discussed in "Understanding the Chicago Homer."
The main settings are quite obvious and let you specify whether the search item occurs more often or less often than a specified number, or between one number and another. Thus a search for words occurring more than 1,000 times will return a list of 32 lemmata, and a search for words occurring only once will return a list of more than 3,000 nonce words, or hapax legomena.
If you look for repeated phrases, the frequency part of the query form shows up with the title "Frequency and Length" and gives you the additional option of specifying the length of the search phrase in words. Thus a search for repeated phrases that are longer than 40 words returns a list with 31 hits.
Fuzzy matching is fully explained in "Understanding the Chicago Homer." You can ignore it here and rely on the default settings. Basically it is a technique for making sure that the computer can bundle repeated phrases that differ only in minor grammatical respects and would be processed by a human reader as "the same."
The "fuzzy matching" section of the query form also includes two additional filters. You can execute a search that will exclude, include, or be restricted to repeated phrases consisting only of function words. For most purposes you will not want to be bothered by such repetitions, and "exclude" is the default setting. But they are informative for some purposes and are more fully discussed in "Understanding the Chicago Homer."
You can also set your search to include, exclude, or be restricted to phrases that contain names. The default setting is "include."
You click on your choices in the word type list box. For the sake of convenience, the major word types of noun, name, adjective, and verb appear at the top of list. The other word types follow them in alphabetical order.
You can select or deselect more than one word type by holding down the control key while clicking on the chosen items separately.
Thus looking for names that appear more than 50 times produces a list of 45 hits. Looking for names or nouns that appear more than 100 times produces a list of 110 hits.
Every inflected word form is defined by some combination of the categories of tense, mood, voice, case, gender, person, and number. You specify choices by selecting the appropriate values in the boxes for the seven inflectional categories. You can select/deselect more than one value from each box by holding down the control key while clicking on a term. If you specify inflectional search criteria, you will probably want to report your results at the word form level.
You can generate lists with different levels of granularity. If you specify aorist and passive in the tense and voice list boxes, the report lists 223 lemmata and 544 word forms. If you add optative from the mood list box as a constraint, you get 19 lemmata and 23 word forms. If you add 1st person, you get 4 word forms.
If you want to look for subjunctive forms of -mi verbs you can type %mi in the search term panel and check subjunctive in the mood box. The return will list 40 lemmata and 111 word forms.
The line range panel allows you to limit a search to text segments you define for particular purposes. It works the same way whether you look for words or repeated phrases.
You specify your text segment by entering information in rows of the line range table, which consists of Text, From, and Through columns. The following is an example of a text segment consisting of two line ranges:
If you want to specify whole books, the application will supply some information automatically. Thus the following example defines a text segment consisting of Iliad 3-5 and Odyssey 8:
Watch out for the implications of Boolean logic in the any/all option. "Any" and "all" correspond to Boolean OR and AND. Assume you are looking for nouns in the text segment defined by Iliad 3-5 and Odyssey 8. If you choose "any," the search will retrieve all the nouns in Iliad 3-5 as well as all the nouns in Odyssey 8. There are 710 such nouns. If you choose "all," the search will retrieve only the 119 nouns that occur in both Iliad 3-5 and Odyssey 8.
The "all" option, or Boolean AND, is a powerful tool for studying shared vocabulary or repetitions. For instance, looking for repeated phrases that occur in Iliad 16 AND Iliad 22 will draw your attention to the ways in which the deaths of Hektor and Patroklos mirror each other.
Narrative, Speech, and Speakers
The final three panels on the search form allow you to limit searches by narratological criteria.
The simplest constraint lets you specify narrative, speech, or either (the default setting). The remaining two constraints allow you to refine a search within speech. Speakers are defined as male/female or mortal/immortal, which allows you, for example, to compare adjectives used by male or female speakers. You can also specify a particular speaker.
A search that specifies only a speaker will generate a list of all words or repeated phrases used by that speaker, whether it is the 92 words used by Thersites or the 954 words used by Penelope.