The word count of Elizabeth Montagu’s letters stands at 183,000. My transcription methods have undergone some changes over the years: most of the earliest transcripts have been checked and re-edited when necessary, but for example the earliest letters to the Duchess of Portland don’t retain line breaks.

Encoding in the letters

The codes in the letters are based on the COCOA system.

  • underlining: (^April^)
  • superscript: y=r=
  • text crossed out: [\Prescripti DELETED\]
  • text inserted above line: (\they/)
  • code-switching from another language: (\je suis\)
  • unclear text: [\me\]
  • emended: a[{re}]

I have used emendations when logical deduction helps determine the word in question, for example when an ink blot covers part of a word and context makes it clear what it is.

Editor’s comments are in all caps. [\BLOT\] and [\TEAR\] accompany unclear cases as additional information. For example: “are you afraid of staking first Miss Carter? Is [\BLOT\] [{that{] the generous & good Miss Carter!”

In addition to deleted text and insertions, self-corrections can be searched with INTO, as in [\you INTO me\].

[\ADDED\], [\IN PENCIL\], [\IN MARGIN\], [\ADDRESS\], [\REST OF LETTER MISSING\] and similar comments point out additions by another hand, layout information, and such issues. These encodings reveal, for example, that Montagu’s letters written in the 1730s don’t have any underlined passages whereas in the 1760s there are over eighty such cases, especially where she quotes from literary sources. Montagu takes up underlining from the 1750s onward as a way of drawing attention to a passage of text (quotations, code-switches).

Header information

Every letter is preceded by a header like this, following the conventions used in the Corpus of Early English Correspondence:


The X line provides the writer’s name in full, and the Q line consists of the following information:

  • A = autograph letter. C = copy would be an alternative, but only autograph letters have been selected.
  • Year.
  • The relationship code between the writer and the recipient.
    • FN = nuclear family, FO = distant family, TC = friends, T = acquaintances
  • The recipient ID. The letter header shows the recipient’s full name, but the IDs can also be found in the Correspondents data and through Search Letters.
  • The writer’s name in short.

The header then gives the full names of the writer and the recipient, the call number of the manuscript letter, the date of writing as closely as possible, and the place where the letter was written. The call number IDs stand for the following collection and archive information:

  • MO: Montagu Collection, Henry E. Huntington Library, San Marino.
  • MS Hyde 7, MS Eng 1531, Eng 1565: Houghton Library, Harvard University
  • BL Add. 40663, Add. MS 42087, 59438 Dropmore papers: British Library


Please note the (perhaps not very intuitive) chronological principle: each decade begins with letters written in the first year of that decade, so that the year 1740 concludes the correspondence of the 1730s, and the 1740s begin with 1741. Don’t look for a letter written in 1740 from the 1740s folder: you won’t find it there.

Manuscript tampering

The later owners of the letters have sometimes left their handprints on them. The letters are often undated, and dates marked in pencil are usually later annotations. And sometimes it’s difficult to tell even from a high-quality image whether a correction has been made by Montagu or by someone else, for example her heir and editor Matthew Montagu. In the image below, preposition stranding (the House which this Grotto is join’d to) has been changed into the normative pied piping construction (the House \to/ which this Grotto is join’d to). The ink and the strikeout suggest that this is a later change, but it isn’t entirely clear. When possible, I’ve noted whether I think the cancellations and emendations are by Montagu or by someone else.

Syntax corrected in letter MO 423, 1747

The next images illustrate the potential gap between a manuscript letter and an edited letter and show how an editor-owner of the letter could prepare the manuscript for printing process.

Here is an image of manuscript letter MO 422 (1747) in which Montagu describes an embarrassing incident during a visit to Mr Sloper’s house where she accidentally inquired after Mrs Sloper’s lover:

Changes made in in letter MO 422, July 6? 1747

Notice the endnotes, the added pronouns, the changes in capitalisation and the revised spelling.

Here is this paragraph as it appears in Matthew Montagu’s edition from 1825 (Archive.org):

Letter MO 422, printed and edited in Montagu ed. 1825: 77

And, finally a section of this letter in the Bluestocking Corpus, edited in an attempt to mark down Matthew Montagu’s changes:

[\a INTO A\] few days ago I [\carryed INTO carried BY MM?\] M=rs= Donnellan, & the little Pere, to see M=r= Sloper’s Gardens & House, at a time when I was assured he was absent on his Election, but seeing a Man ride up the Avenue at the same time, [\I BY MM?/] took it into my head it might be M=r= Sloper, so [\I BY MM?/] did not alight immediately. [\the INTO The\] Housekeeper came to me, & ask’d if I w=d= walk in, I said I should be glad to see the House if (^M=r= Cibber UNDERLINED BY MM?^) was not at home, the Housekeeper look’d as [\agast INTO aghast WITH [\h/] BY MM?\] as if she had spoild a Custard, or broke a jelly glass; I coloured, M=rs= Donnellan twitterd, D=r= Courayer sputterd half french half English, & began to search for y=e= Case of a spying glass I had dropt in my fright. As my organs of speech, rather than of sight, seem’d defective, I was little interested for my perspective, but sat in the Coach making melancholly reflections on my mistake.

The outcome tends to be messy when you flag editorial intrusion, but I’ve done this in order to be as transparent as possible in my reading of these letters.