One of my favorite classes to teach is History of the English Language, and ever since I first started teaching it as Carol Percy’s TA at Toronto, I’ve been using an assignment geared toward getting students comfortable with using Old English dictionaries, even if, as is usually they case, they don’t use Old English. At Toronto, we had the benefit of a subscription to the Dictionary of Old English, but at my current university, New Mexico Highlands, we don’t. But fortunately, the DOE has free accounts for individuals to access the material up to 20 times, which I find is generally sufficient for almost all student needs in HEL. This semester I did a thorough revision of my handout introducing students to dictionary resources, and I thought I would post it up in case others wanted to make use of it in their own teaching, whether in a classroom or self-directed. Feel free to share favorite search techniques as well. I tend to go through it in class up on a screen with students, with plenty of side excursions based on their input (this semester we got distracted by compounds for teeth). It’s easy to spend anywhere from an hour to two hours on something like this, and I tend to follow it up with an essay using the resources to analyze select compounds in Old English poetry (I’ve used Beowulf, Maldon, and Brunanburh in the past).
Old English Lexicography: Online Resources
There are three main dictionaries available to you in some form online, and two of the three are also available in hard copy. Unlike later periods, the OED is of only limited use in studying Old English, because it does not include any words that do not survive from Old English in later periods, and because it is organized based on modern, rather than medieval, spelling.
The three main dictionaries available to you are:
A Concise Anglo-Saxon Dictionary by John R. Clark Hall (Clark Hall)
An Anglo-Saxon Dictionary edited by Joseph Bosworth and T. Northcote Toller (Bosworth Toller)
The Dictionary of Old English, A-H, edited by Angus Cameron, Ashley Crandell Amos, Antonette diPaolo Healey, et al. (DOE)
Clark Hall and Bosworth-Toller are both available in hard-copy in the reference section in NMHU’s Donnelly Library. I will also put my copy of Clark Hall above the mailboxes in Prescilla’s office, and you are welcome to make short term use of it in the building, although I would ask that you keep it here and return it when you are done. (For people on the internet, obviously you can’t come in to the department secretary’s office to consult these dictionaries, but I would guess that many university libraries have copies of both Clark Hall and Bosworth-Toller in their reference sections).
Each dictionary has different strengths and is suited for different tasks.
Clark Hall only provides very brief entries, and is most useful as a quick reference while translating.
Bosworth-Toller provides much fuller entries, including example quotations, but it also suffers from some of the eccentricities of its nineteenth-century editors, perhaps most clearly seen in the entry for “Beowulf”, in which Bosworth speculates at some length about the existence of an undiscovered saga about Beowulf mouldering in a Swedish library. Bosworth-Toller is the dictionary equivalent of an eccentric uncle.
The DOE contains the most comprehensive entries for words and is the gold standard for lexicographic research in the field. But the DOE project, which began in 1970, is still ongoing, and at present has only published entries up to the letter H.
Using Clark Hall
Online access to Clark Hall is provided by the Germanic Lexicon Project at Penn, which hosts images of the pages of the hard copy of the dictionary. Each image is referenced based on the first and last headwords on the page.
Thus, the first image is of the page containing all the entries from a to ablisian. Clark Hall (and other dictionaries) alphabetize special characters as follows: æ comes between ad and af, and þ/ð come immediately after entries for t. OE dictionaries typically choose to consistently use one character or the other of þ/ð consistently in all headwords, although this can lead to an amusing situation in which sometimes headwords are given with spellings that never actually occur in extant Old English texts.
Because you have to click through multiple pages, rather than riffling through them like you would the hard copy, the online scans of Clark Hall can in practice be a little unwieldy, especially, if you end up having to flip pages due to spelling variation. But Clark Hall is still the most convenient dictionary if you just want to know what a word means with as little fuss as possible.
Here is a sampling of entries. They typically give the bare minimum of information. The abbreviations to the right indicate which texts the word may be found in. There is a list of signs and abbreviations provided at the beginning of the dictionary, which you can use if you want to know what particular abbreviations mean. The entry for goldsmið, for instance, contains an Æ, indicating that this word is found in the works of the author Ælfric.
The online Bosworth-Toller is a little more user friendly than Clark Hall, as it actually allows you to search the dictionary. What’s more, the search box actually has an auto-complete feature. This is very useful when looking up Old English words. When you encounter an Old English word in the wild, it will often have an inflectional ending which won’t be present in a dictionary headword. Thus, it can be useful to focus on entering the first syllables of a word and using the auto-complete feature to let you see which options possibly best reflect your word.
Another tricky thing about searching for Old English headwords is that, while the consonants in Old English tend to be relatively stable, the same cannot be said for the vowels, which may change due to mutations or in strong verbs or due to the dialect of the source. If you have difficulty locating a word with the auto-complete option in a simple search, it may help to try the advanced search option.
Sometimes doing a headword search with the “contains” or “starts with” options will be the ticket for finding a word that is eluding simpler searches. The “contains” option is also a very useful way for identifying morphological elements. For instance, doing a “contains” search for gold will bring up all the entries that contain that element. This can be very useful when looking for compounds in Old English.
The “find entry that:” search box can be useful if you want to try out a variant spelling, but its downfall is that it may pull up entries that don’t pertain to the word at all, but merely contain it as part of one of its example quotations. Judicious use of this box can be helpful, but also frustrating, as it will typically bring back way more results than are relevant to your search. One way it can be useful, however, is as a way of doing reverse look-ups based on a modern English word, although again, depending on the word you may end up getting more than you expected.
Sometimes this process can be a little bit tricky. If you wanted to know the Old English word(s) for shit, for instance, you wouldn’t find any entries unless you played around with synonyms a bit. A search for poop leads to the etymologically unrelated nautical word for poop-deck (from the Latin puppis), and bowel movement also gives no results. Thinking like an eccentric 19th century lexicographer turns out to be challenging at times. But excrement is a far more rewarding search term. Dung, while a common term in the definitions, actually turns out to be a bad search term because of the prevalence of the –ung suffix in Old English words.
Entries in Bosworth-Toller provide more information than those in Clark Hall. The entry for goldsmiþ (notice that Bosworth-Toller uses þ instead of ð), for instance, looks like this:
Here we get illustrative quotations and a much fuller indication of potential sources, although again, interpreting the sources often requires more advanced knowledge of the field. These fuller entries mean that Bosworth-Toller is a much more useful dictionary for research purposes, and is especially useful for headwords that have not yet been published by the DOE.
Creating an Account to Access the DOE
The DOE is a subscription based service, and unfortunately I don’t think I can convince the library to get an institutional subscription for the benefit of this one class. But, fortunately for us, the DOE allows you to have free access up to 20 times per year, which should be plenty for the assignments we will be working on in this class. There is an FAQ that explains how to sign-up for free access to the DOE on the main page (http://www.doe.utoronto.ca/pages/free.html), but I will also provide screenshots of the whole process.
The first step is to go to the DOE Project Online Store and click Login. Make sure that you don’t click on any of the options to purchase a subscription.
On the Login page, select the option for New Customer
The next page will require you to enter a bunch of information about yourself. I can say from personal experience that the DOE is not going to sign you up for a bunch of mailing lists or anything of the sort. No unwanted e-mails about increasing the size of your lexicon.
Once you have filled out the required information, you will be sent to a confirmation page, at which point you are ready to begin accessing the DOE.
Using the DOE
The reason the DOE is such a time-consuming project is that it is attempting to be a comprehensive dictionary of the language. Before they even began writing entries, they created a searchable corpus of all texts written in the period (sometimes including multiple manuscript variants for key texts). The DOE takes account of all known instances of a word being used in extant sources and includes information about the number of occurrences, attested spellings, and provides much fuller examples of illustrative quotations. Looking at the entry for gold-smiþ demonstrates what this looks like in action. The DOE also includes this as a sample entry accessible from the main page. This sample entry can be very useful to consult if you don’t remember what the parts of an entry are. However, the links on the sample entry are non-functional. This entry looks like this:
Noun, m., cl. 1
Att. sp.: goldsmiþ, goldsmið, golsmiþ (ÆColl) || goldsmiðes || goldsmiþe || goldsmiðas; goldsmiðes (nom.pl., Rec 20.1)
Met 10.33: hwær sint nu þæs wisan Welandes ban, þæs goldsmiðes, þe wæs geo mærost? (Bo 19.46.16 þæs foremeran & þæs wisan goldsmiðes ban Welondes; cf. BOETH. Cons.Phil.metr. 2.7.15 ubi nunc fidelis ossa Fabricii manent?).
ÆCHom I, 4 209.92: ealle þas goldsmiðas secgað þæt hi næfre ær swa clæne gold ne swa read ne gesawon (cf. Vit.Iohan. 2.57.2 omnium aurificum <officinas> circuiuimus, et omnes dixerunt tam purum … se nunquam uidisse aurum).
Gen 4.22: be Sellan he gestrynde Tubalcain, se wæs ægðer ge goldsmið ge irensmið (L slecgwirhta & smið on eallum weorcum <æres & ysenes> [MS ærest of ysene]; cf. Gn: malleator et faber in cuncta opera aeris et ferri).
Ch 1497 50: and freoge mon Man hire goldsmið, & his eldestan sunu.
Ch 543 1.1: þis his þas anes hiwisces boc at Winterburnan, & ðas oþres hiwisces on Whit þe þærto hyrþ, þe Eadred cyng gebocode Ælfsige his goldsmiþe on eche hyfte.
Rec 9.3 1: her is on þysse Crystesbec siu geswytelung þære healfre hyde æt Pottune þe Ælfhelm Leofsige sealde hys goldsmiþe.
Rec 20.1 1: Ælfric and Wulfwine Eadgife goldsmiðes geafen to broþerrædenne twegen orn weghenes goldes þæt is on þis ilce boc her foruten gewired.
ÆColl 205: habeo fabros, ferrarios, aurificem, argentarium, ęrarium, lignarium et multos alios uariarum artium operatores ic hæbbe smiþas, <isensmiþas>, goldsmiþ, seoloforsmiþ, arsmiþ, treowwyrhtan & manegra oþre mistlicra cræfta biggenceras.
ÆGl 301.15: aurifex goldsmið.
ClGl 2 515: aurifex goldsmiþ.
Fort 72: sumum wundorgiefe þurh goldsmiþe gearwad <weorþað>; ful oft he gehyrdeð ond gehyrsteð wel, brytencyninges beorn (‘for some marvellous gifts are prepared by the goldsmith’; the word has alternatively been taken as an otherwise unattested compound *goldsmiþu or *goldsmīþ, synonymous with ON gullsmíþ ‘goldsmithery, goldsmith’s art / work’).
- glossing ferrarius‘blacksmith’ as if aurifex‘goldsmith’
ÆColl 231: ferrarius respondit se golsmiþ andwyrt (cf. ÆColl 205 quoted above).
Lat. equiv. in MS: aurifex
See also: gold, smiþ
The main downfall of the DOE from a student’s perspective is that illustrative quotations aren’t translated. But if you need assistance working through some of the examples for your essay, please let me know and I will be happy to help.
Before you do any searches, make sure that you login. In the upper right hand corner on the main search page, there is a small box that will either say User (if you haven’t logged in) or your name (if you have).
The DOE is searchable in a range of ways. Most obviously, you can search by headword. However, there are a large number of ways to search that can help you in your research. The search function is available once you have entered the main dictionary. It is located immediately above the section entitled “Short Titles and Bibliography,” which is another useful section if you do not know the Short Title for a given work.
Many OE words are difficult to look up because the inflected form of a word does not necessarily have a similar spelling, especially with regards to vowels, as the base form of the word. If you have trouble locating a word in the DOE (up to H), simply search under the field Attested Spelling. For example, if you were reading the gloss to the Lindisfarne Gospels and came upon the word bulgon, you might be confused. But a search of the DOE by attested spelling shows that it appears under the headword belgan. This will almost always be the easiest way to look up a word that appears in a text you are translating, and it is in my mind the thing that makes the DOE incredibly convenient.
My favorite kind of search in the DOE is the Occurrence field, which you can use to search for how often words appear. However, when using it keep in mind (as always) that it only searches up to H. However, the Occurrence field is even more powerful when used in conjunction with another field: Citation Reference. To use multiple fields click the More Fields button on the search page. In Occurrence type the number of occurrences you are interested in and give the short title of the work in the Citation Reference field. As an example, let’s use Beowulf, whose short title is Beo.
A search for 2 occurrences where at least one is in Beo returns 199 results under 138 entries (the extra results are due to words that occur twice within the same poem). Let’s take a look at the second result, ān-pæþ. This word, which means “a path for one person, narrow path, defile” occurs in two locations, Exodus (short title Ex) and Beowulf. We see that not only do the two poems share the word, but they share an entire phrase, “enge anpaðas, uncuð gelad.” Using the occurrence field in this way can help reveal contextual connections between works. Words that have only a single occurrence are a special case, and are referred to as hapax legomenon (sg., plural is hapax legomena or informally hapaxes). These words are actually quite common in OE, both due to the limited number of extant manuscripts and due to the frequency of compounding as a stylistic device in OE verse. The term hapax legomenon is fairly common in studies of ancient languages. For instance, biblical scholars might identify hapax legomena that only appear once in the entire Bible.
Using the DOE Web Corpus
Even though the DOE only goes up to H, the body of material on which it is based is all available online and you are encouraged to make use of it as well in your research. The Web Corpus is also available from the DOE front page. Just click on the link for Subscribers underneath the Web Corpus. When searching the web corpus, there are a couple of important things to keep in mind. The first is that special characters are entered through the use of capitals. A=æ, D=ð, and, T=þ. Also keep in mind that the web corpus does not recognize variant spellings on its own. Consequently it is generally useful to leave out case-endings, which would needlessly eliminate valuable results. Because it does not recognize variants, it may be useful to perform multiple searches using different variants. If the word you are searching for comes before H, use the attested spellings from the DOE to find all the citations for a word. If the word does not come before H, use the variants listed in Bosworth-Toller as a starting place, and on your own perform multiple searches using common variants. For instance, in words that contain eth or thorn, it will be important to search for both variants.
Searching the web corpus can show how words appear in different contexts. For example, the last word of Beowulf is lofgeornost, which is generally translated as most eager for praise. A Web Corpus search for lofgeorn reveals 12 instances of the word.
Each result gives the short title as well as a letter/number combination. This is referred to as the Cameron number (named for Angus Cameron, the first editor of the DOE). Cameron numbers provide quick identification of texts, and the letter corresponds to the type of text, as follows:
C: Interlinear glosses
E: Runic inscriptions
F: Non-runic Inscriptions
If you need more information about a text, clicking on the blue link will pull up the full bibliographic information.
We know based on the Cameron numbers that Beowulf is the only poetic source. All the other instances of lofgeorn are Bs, that is, prose. Furthermore, if we make reference to the short title list available within the DOE, we can determine what the other texts are.
For instance, the second entry is:
ÆLS (Memory of Saints) B1.3.17
- [0081 (300)] Se seofoða leahter is iactantia gecweden, þæt is ydelgylp on ængliscre spræce þæt is ðonne se man bið lofgeorn and mid licetunge færð, and deð for gylpe gif he hwæt dælan wile, and bið þonne se hlisa his edlean ðære dæde and his wite andbidað on ðære toweardan worulde.
The short title ÆLS (Memory of Saints) corresponds to Memory of Saints in Ælfric’s Lives of Saints. The prose texts given by our search are all saints’ lives or homilies or translations of the Benedictine Rule. A partial translation is: “The seventh sin is called iactantia, that is idle-boasting in English, that is when someone is eager for praise…
We can see here that although Beowulf is described as someone exemplary within a pagan tradition as most eager for praise, there is a double meaning in the Christian context of the poem, so that the audience sees Beowulf as aspiring to pre-Christian ideals that are ultimately at odds with Christian conceptions of morality.
As an example from another poem, we could choose the compound wælstow from line 95 of The Battle of Maldon, which means battlefield, literally, slaughter-place. The line reads “god ana wat / hwa þære wælstowe wealdan mote” (god alone knows who may control the battle-field).
To search the DOE Corpus, enter wAlstow into the search field, leaving off the inflectional ending to get the greatest number of hits. The Corpus search reveals 45 hits. Looking up the short titles tells us that the word occurs primarily in poetry and chronicle texts. Furthermore, even without knowing much OE, we can see that wælstow tends to occur in conjunction with words related to wealdan, to control or wield. The emphasis placed on alliteration in the lines serves to heighten the connection between battlefield and control. One way you could go with your essays would be to search for other connections between words.
In searching the Web Corpus, it may be useful to limit your searches to poetry (although as seen by the previous example, restricting your searches could cause you to miss other details). In order to limit searches to verse, simply enter a in the field “Restrict to one of more works using Cameron number.” Make sure to use a lower case a, as the site will interpret an upper case letter as an ash character, and you won’t get any results.