Australian Newspaper text correction guidelines
Dear text correctors
Ann Manley one fo the top text correctors has kindly organised the creation of guidelines for text correction to help new and current text correctors. The guidelines have been reviewed and edited by the top 5 text correctors (Julie, John, Fay, Maurie and Lyn) and myself. If you have amendments or additions you would like to suggest for the guidelines please discuss them here with other text correctors. I will leave the guidelines here in the forum for review for a few weeks before making any final revisions and putting them into the 'Help' section of Trove/Australian Newspapers. Many new users have asked me for guidelines on how to correct text, so I am hoping this will be a useful guideline for the text correctors community. There are several thousand people correcting text. Your input is of great value to us.
Last edited by mraadgev [NLA]; 14-02-2011 at 03:59 PM.
Many thanks to you Rose and Julie, John, Fay, Maurie and Lyn (and Ann?),
I'm very glad to see this clarification appear (and keep me on the straight and narrow).
I am relieved to say that although I have done only a little text-correction, what I have done is consistent (I think and dearly hope) with Ann's excellent advice. There are only two things that occur to me reading the existing Guidelines. I wonder if we need to distinguish between:
(a) keyboard-specific instructions (of which there are a few in these guidelines) for Windows-based computers and Macs (which are slightly different); and
(b) possible confusion between EM rules (—) and EN rules (–). I think the Guidelines talk only of em rules but include actual examples only of en rules. To get to the point, mostly what we seem to come across in the newspapers (in birth/marriage/death lists, for example) are em rules (the REALLY long dash, also used as a rough equivalent to parentheses—like this—on occasions). Modern style guides call for en rules (longer than a hyphen (-) but shorter than an em rule (—) in spans of dates (like 1998–99) or in a spaced form as an alternative to em rules – like this – but I cannot recall an obvious use within the newspapers in the corrections I have made. Nevertheless, we may need to be alert to the three different lengths of dash. On a Mac, an em rule is 'option-shift hyphen'; an en rule is 'option hyphen'. On a Windows machine, an em rule can be got by holding down the ALT key and typing 0151; an en rule by holding down the ALT key and typing 0150. (There may be other ways of doing it.)
Sorry if this seems a bit obsessively detailed!
Last edited by David.Rowlands; 15-07-2010 at 02:59 PM.
Great guidelines! And as an ex-editor, I sympathise with David's comment about the differences between em and en dashes.
A couple of minor thoughts about the guidelines:
1. In the bullet point that begins "If text is obscure...", in addition to recommending searching other Australian newspaper articles, perhaps include a note about using Google to find what the text behind the blotches may actually be. In particular, since many of the 19th century articles were duplicated in publications across the world, I've found Google Books a good way of locating alternative copies of articles and even family notices (in publications like the Gentleman's Magazine, for example). It's then possible to use these alternative copies of articles to guide text corrections, comments or tags.
2. Related to the bullet point beginning "Whilst no abuse...", is it possible to recommend a standard phrase/sentence to use in comments to report abuse? Looking towards the future, if there's ever an automated way to report abusive activities, it might be easier to harvest abuse reports made through the current system of adding a comment about the abuse by searching on a standard phrase/sentence. Otherwise, "abuse" may refer to the content of the article ("abuse of power", for example) rather than abusive use of the text correction, tagging or commenting functions.
The font used in the draft correction guidelines (and perhaps even in this forum) may seem a bit bizarre re comments regarding the M‘ names and specifically to the part:
"M‘ surnames interpreted as M’ = Mc" (hope the differentiation is more apparent here)
(in the pdf draft guidelines they appear to say the same thing, i.e. repetition of M’ or M')
Close observation in newspapers will reveal that the apostrophe in these names in most cases could be regarded as a backwards apostrophe (looks like the start of a little 6), rather than a forward apostrophe. [Am sure there is probably correct terminology for the different apostrophes, but I am not a linguist].
Maybe one of the former correspondents above may be able to expand further for the clarification of others.
Further, have used the description "Em dash" in the guidelines as that is how the Special Character in the pull-down menu is described.
Also, the guidelines were produced without any real knowledge of Mac computers; so again, again further comments regarding variations that might apply to same would be greatly appreciated.
Consider that it ought to be Recommended reading for those using Trove. It provides a coverage of techniques, problems and solutions en counted in attempting to correct text from the newspaper ETT project.
It has certainly clarified a number of points such as the limitation on the number of spaces between characters- 'what you want to see is not what you get'. The fact that the original document (as scanned) remains on record is also an important reassurance for those attempting to edit.
I am having a problem with Tabulated data as is frequently found in Advertisements and published lists. The scanning can result in columns being split into separate lines – this can be fixed by entering the transposed data onto the line & subsequently deleting the extra line – at least the context is consistent but the alignment of the data in the columns suffers in appearance (the limitation of number of blanks but .. could be substituted). What happens when there is more than one column in the tabulation is unknown – the actual heading might be a large number of lines previously edited. The lines (rows) in the tabulation do not appear to align.
What should one do with 'Stock Brands'?, as used on horses & cattle, these can involve Letters & other symbols arranged in a variety of ways sometimes over up to 3 lines. The skill of the type setters for these early records was truly amazing!. I have been guilty of attempting to describe the brand in words but it would appear better to leave the ETT as is, even if this results in extra lines which are mean less, no longer aligned with the other characters. I would appreciate advice on what is a consistent approach to this data – to do nothing appears to create a bit of a mess.
It has also been assumed that preservation of format variation such as the use of Old English is not required but a translation is in order. Also the preservation of Bold type & Capitalisation of Text is not mandatory and space within words for emphasis is also not important since it would make subsequent search impossible.
Most of my editing has been on records from 1840's & 1850's. It makes one aware of the evolution of language which is one of the byways or diversions from the primary research objective.
Hi, thanks very much to those who have been involved in creating the guidelines. It has forced me to move over to Trove so I can use the dropdown list of symbols (a brilliant idea).
I've got to admit I still have been using M' instead of Mc - keeping in mind that names with apostrophes like D'Arcy, O'Brien and M'Donald are still searchable by inputting the search term darcy, obrien or mdonald (or alternatively arcy, brien, donald although this can mean wading through a great deal more irrelevant hits). Even if all M' are changed to Mc to make them easier to find, users still need to know how to search for other types of names that contain an apostrophe.
Also if anyone comes across my corrections where the correct letter and incorrect letter are still present, my sincerest apologies, my delete button was no longer as sensitive as it used to be and required more pressure to respond. Should have been paying more attention to the corrected text...
In response to some recent comments by various text correctors:
We don’t need to know when you’ve corrected text, as this evident by the history associated with corrections for each article; nor do we need to know about minor problems with the text.
Sometimes articles are spread over more than one page – if hovering over the relevant text does not allow you to edit, you may need to scroll up or down, to find the relevant page (which will appear in blue type) and click on this page number, before you can edit text for that page. You need to save text corrections on a page by page basis, before navigating to another page.
Correctors often state that parts of articles are missing, when in fact they are incorporated into another article on the same page; please check other articles on the same page before commenting; otherwise follow the guidelines for missing text where possible.
Last edited by annmanley; 02-08-2010 at 06:50 PM.
In regards to the M' - I understand that the ' character used in newspapers is typographically termed a superscript 'turned comma', but I have not noticed a computer font which has this character. I don't know if there is a linguistics term for it, but I was told many years ago that M' represented the phonetic Mac, and that the ' was not a superscript c as some non-Irish persons had thought.
This being the case, changing M' to Mc will be incorrect when we are dealing with names such as John M'Arthur, who evolved from writing his name that way to MacArthur (and towards the end of his life Macarthur, which his descendants seem to have favoured). Test searches using MArthur (since system does not recognise ' for search purposes), Mcarthur, and MacArthur give different numbers of matches for each. By changing the ETT text where the turned comma has been used in the newspaper, will we be making things easier for persons searching, or are we going to cause some confusion? Whatever is decided, I think it is important that it be indicated in the on-line Help as well.
I agree with tagging with a full name, when it is definitely known, especially where a surname consists of a word existing in the English language, or where the system's 'fuzzy' search engine matches on another word, and returns many spurious matches. For example, I am editing (slowly) articles on members of the LOANE family; a full search on 'loane' returns 904,702 matches, (because of 'loan' being a common word), whereas some judicious searching using full names, initials, and prefixes such as Mr, Dr, Major, etc., leads me to believe that there are only about 3,000 true matches. Perhaps the guidelines could place more emphasise on the value of tagging in these circumstances.
A similar thing may occur with the names of 'ships' (I use the term generically). For example my ancestor owned the 'Derwent' schooner, but a search on Derwent throws up many spurious matches. (I hate to think what 'Martha' or 'Sally' would turn up.) Could I suggest we standardise on the method of tagging these by using the prefix 'ship' before the name of the vessel, without using the particular designation of schooner, brig, etc.? (Which many researchers might not be aware of in any case.)
I have found it beneficial, especially when correcting a lengthy article, to copy the corrected text into a word processor file, and then running the spell checker. This will show up any of the common letter/number substitutions which can be hard to pick up during editing, and using CTRL + F allows you to quickly find the errors and make the final corrections to the ETT. While not a 'standard' to be conformed to, perhaps this could be listed as a tip in the guidelines. It works very well with a wide screen using side-by-side windows.
I have come across some pages where there is a line of text missing, both in the ETT and in the image, caused by a crease in the original newspaper when it was copied. What is the best way to indicate this fact? In a comment? Is there some way of marking the point at which this occurs?
Hello Ann, I started using the private tags "corrected" and "working on" to indicate articles that were completely corrected or that I intended to come back and finish. I often don't get time to correct a whole article but I like to try so have been using these tags for myself.
Originally Posted by annmanley
I have recently been accepting the default of "public" when I use the tag "corrected" for articles I consider fully corrected. I assume this helps others - is this a problem?
It would be good if there was a function to mark articles that were fully corrected and then a second step to mark those verified by another user.