Towards a new platform for texts

Currently the Oxford Text Archive offers a repository from which texts can be downloaded, and the Electronic Enlightenment offers a platform for exploring and reading historical correspondence. This post explores some of the ways in which we are experimenting with new ways to explore these texts, moving towards a new common platform for textual resources in the Bodleian Libraries.

Searching for Rhetorical Figures

We might recognize rhetorical figures when we read them, but how do we cross-search the collections looking for examples? We carried out some experiments by putting collections of EE and OTA texts into a variety of corpus search engines. This threw up examples both well-known and unexpected.


Searching for sequences of at least four words starting with the same letter reveals dozens of examples. e.g.:

[ota_ee_english:1636/descreCU0030050b1c] A bow bends back because, when the shape of…[ota_ee_english:1791/burkedOU0010304a1c] … all others the most horrid; that of betraying by being betrayed.

The second is an example of chiasmus (inverting the order of repeated words or phrases) as well as aliteration. Finding letters with examples of one sort of rhetorical figure is likely to reveal a rich source of other figures. The letter at, for example also features more alliterations and various types of repetition. e.g. “an academy of cabal and confusion”, ” It is dangerous for you fully to trust those by whom you are not fully trusted”, “fear of the French faction”.

A similar rich seam was found with the following letter from Hester Lynch Piozzi:

[1798/piozheUD0020481a1c] … seems to have a Feeling of Fear, however high his heart may beat with Indignation.

And this letter also contained a number of parallelisms and repetitions of various sorts, e.g.:

“Another Source of Comfort which opened upon us as we drove along was the encouraging Sight of universal Loyalty, and Ardour in the honourable Cause of defending ourselves against French Threatenings, French Plunder and French Principles.”)

Some further examples of alliteration:

[1815/piozheUD0050400a1c] … teaching Children are our boast, yet we have hard hearts here too; and little Tradesmen…

[1698/fitzwiVH0010364a1c] … & some a better price, my own stemmed sweet scented so far beyond, that I…

[1802/piozheUD0030375a1c] Let not a Heart so sensible so sincere want that Support from within…

The last example also including the following:

“Sick or well, sorry or glad; nobody Sure does write such Letters as our dear Mrs. Pennington — it is because nobody else writes from the Heart.— I suppose. Let not a Heart so sensible so sincere want that Support from within which we are told the Consciousness of Rectitude will always bestow; and let no ill Thoughts steal in to destroy the Operation of virtuous Talents and Conduct.”

[1724/popealOU0020236a1c] … to be turnd into a Line of Wilderness with wild winding walks for the convenience of passing…

[1816/bentjeOU0080506b1c] … haveable for taking first measurements: and then what with weather what with languor and inaptitude,…

So, in summary, while alliteration in and of itself might not be of primary interest, it offers a way into exploring the texts.

Further explorations of rhetorical figures

The tools also allow searching by part-of-speech, for example for repeated instances of one particular word class. In this example, William Cowper, uses 4 consecutive adjectives:

” have received but one Visit since I came — I don’t mean that I have refused any, but that only one has been offered. This was from my Woollen Drapier,3 a very Healthy Wealthy Sensible Sponsible Man, and extremely Civil. He has a Cold Bath, and has promised me a Key of it, which I shall probably make use of in the Winter. He has undertaken too to get me the St. James’s Chronicle4 3 times a Week, & to shew me Hinchinbrook House,5 and to do every Service for me in his Power; so that I did not exceed the Truth you see when I spoke of his Civility.” (

And by the unknown “S.L.”:

“Why soe sullen why soe silent What not a lettre not a word in six long weary wett windy weeks Wher’s the fault?” (

sermocinatio / hypophora (“the speaker answers the remarks or questions of a pretended interlocutor”, or simply “asking questions and answering them”)

“Now you might say — Why are you against it ? ” and “Were I to say as much Mr White might be in a rage, and Mr Long might say — Mr B. why will you propose such things? — You see it can not be done. The consequence would be, that between Mr White, and Mr Harrison, and Mr Long, possession never would be obtained.” bentjeOU0060303b1c (Bentham, 1800)

“Even in and about London with great deference to the better judgements of Messrs. Pitt and Dundas I cannot bring myself to believe that those whose opinions I value most could reconcile to themselves my signing. My enemies might say, and with propriety too, he has by giving his signature proved Mr. B. ‘not to have been incorrect in his charges.’ In short, exclusive of all this the internal monitor conscience precludes me from signing and this with me must ever be superior to all human opinion.” scotdaRH0020249a1c (David Scott, 1800)


“You must excuse me for giving you a Line of Latin now and then since I find my self in some danger of Losing the Tongue , for I perceive a new Language , like a New Mistress , is apt to make a man forget all his old ones . ” addijoOU0010010a1c (Addison 1699)

One way to explore simile is by searching for the pattern ‘like a’. Not only can concordances can be generated to find examples, but the tools can also be asked to count instances of repeated patterns, as below:

These patterns can be further explored, e.g. by clicking on ‘like a Philosopher’ above to see concordances of the examples:

and ‘like a mistress’:

and compared to ‘like a Lady’:

and ‘like a Man’:

Grammatical structures in French

The above examples focus on a corpus of all of the letters in EE in English. We also looked at letters in French.

Coordinated adjectives before and after the noun

One brief investigation was to explore to what extent certain forms of French grammar were present in the letters. This example is of a form current in earlier forms of the language but which we would expect to be becoming less widely used and possibly sounding archaic in the eighteenth century. We found it was much more likely to be used with ‘homme’ as the subject than with other words.

[1690/lecljeLO0020039a1c.ddc] … dans paris jusqu’a present composée des plus honnestes gens et habiles de paris: elle est…

[1711/lecljeLO0030359a1c]… santé, et qu’il vous donne une longue vie et heureuse, pour l’avantage de…

[1736/voltfrVF0870332a1c]…a pu avoir une ferme envie d’être honnête homme et sage.

[1744/voltfrVF0930113a1c]… unique, elle était adorée d’un mari honnête homme et aimable.

[1757/voltfrVF1010450b1c] Le conseil de guerre l’a déclaré brave homme et fidèle.

[1763/rousjeVF0170062a1c]… quoique Ministre, ne laissoit pas d’être honnête homme et bien intentionné.

[1760/voltfrVF1060298a1c]… vous pas aussi qu’il vaut mieux être honnête homme et aimable, qu’hypocrite et insolent..

[1693/lockjoOU0040689a1c] Le Sieur Smith passe ici pour un honête homme et fidele dans le commerce, et…

Terms in the history of science

We also combined letters (from EE) and works (from the OTA) of certain authors, in order to cross-search, and compare and contrast usage in these text types (i.e. compare their writing in letters versus published works). The following example is a search in the letters and works of Robert Boyle.

We can see the different word forms matching this wild card search term (showing that it would be difficult to search for them one by one and to see all the results side by side):

And we can explore collocates (words co-occurring in the proximity of the ‘volatile’ words more frequently than you would expect by chance):

We can examine the distribution of the occurrences of the ‘volatile’ words, for example by decade, or comparing letters and works (clear picture from the latter!):

Search term ‘volatil*’ in corpus ‘boyle’. Based on classification: Decade of creation
Category [↓]Words in categoryHits in categoryDispersion
(no. texts with 1+ hits)
Frequency [↓]
per million words in category
1600_925,429171 out of 1668.53
1630_947300 out of 30.00
1640_922,39200 out of 390.00
1650_957,31111 out of 2717.45
1660_91,008,1051848 out of 93182.52
1670_9515,6943336 out of 46645.73
1680_9512,54816510 out of 60321.92
1690_9753,9951788 out of 15236.08
Total:2,958,35487834 out of 285296.79
Search term ‘volatil*’ in corpus ‘boyle’. Based on classification: Type of text
Category [↓]Words in categoryHits in categoryDispersion
(no. texts with 1+ hits)
Frequency [↓]
per million words in category
Letters137,37132 out of 23621.84
Works2,820,98387532 out of 49310.18
Total:2,958,35487834 out of 285296.79

The ‘DiaCollo’ tool from the Berlin Brandenburg Academy of Sciences also allows the visualization collocations over time. Here it has calculated the significant collocates of the word ‘philosophy’ in Boyle’s letters and works at different time periods, and shows them as word clouds in a time sequence. Below is a screenshot of that time sequence from the 1660s (suggesting that this is when the term ‘experimental philosopher’ first comes to the fore):

Exploring a word or concept: la guerre

The use of CQPweb allowed us to search across all of the letters in French for a particular word or phrase, and to explore the results in various ways. First a concordance for guerre:

This can be sorted alphabetically on the right co-text to reveal patterns of modification, e.g. guerre de X:

Having decided that this looks like an interesting way to investigate historical topics in the letters we can run the search again restricting the search to patterns of ‘guerre de or guerre d’:

Looking at collocates for guerre we can explore other patterns:

And crucially, the tools always allow the user to drill down and read the text and its context, by reading concordances, metadata and links to the full texts themselves.

In the DiaCollo tool, we can geta different visualizaton of the words co-occurring with guerre. The screenshot below shows evidence of (among other things) references to the ancient Greek fable Batrachomyomachia, also known as la guerre des grenouilles et des rats.

With these experiments, we are starting to see what is possible when annotations are added to the texts, and when more powerful software tools are added to the repositories. The quest is ongoing!

Search OpenEdition Search

You will be redirected to OpenEdition Search