- How to edit text
- Why text needs editing
- Find articles to edit
- Editing guidelines
- Keep track of edits
One of the great features of Trove is being able to correct the text transcript of the items. This can be helpful when the original scanned image has printed text that isn't very clear or legible. The text column is located to the left of the item viewing pane.
While viewing newspapers, gazettes, magazines, newsletters and books, you may notice that the text transcript doesn't always match the text in the item. You have the power to fix this by editing the transcript to match the image text.
- Go to the image you want to edit.
- While viewing the image, select the green Match text button at the top of the text column, or hover your cursor over a sentence that you want to edit and select the icon.
- If you are not logged in to your Trove account, you will be asked if you want to Login, Sign Up or Make an anonymous correction. Note: You don’t need a Trove account to edit, but if you have one, you can track and access your previous edits when you are logged in.
- Edit the text in the column while following the guidelines on this Help page. Select a sentence, and a green border will appear around it with a cursor to enable editing.
- Save changes as you go by selecting the green Save button.
- When you’ve finished editing and want to save your changes, select Save & Exit.
- To discard any changes made since your last save, select Cancel and exit editing mode.
Why text needs editing
The transcripts are automatically created by a computer when newspapers, gazettes, magazines, newsletters or books are scanned and digitised. Because of this, the text is not always correct, especially if the original source was blurry, faint or in otherwise poor condition.
When you do searches for keywords and phrases in Trove, the search engine checks the original article text AND the transcript to try to find a match. By editing a transcript to match an image, you’re improving the chances of keywords in the item being found.
Editing text to match articles = making Trove searches more successful for everyone.
Transcripts can also be downloaded, so by editing a transcript to match an item, you’re also editing the downloaded transcript.
Find articles to edit
If you’re keen to become a Voluntrove and start editing newspapers, gazettes, magazines, newsletters or books, we recommend finding items on topics that interest you, such as your hobbies or family history.
If you want to find articles that haven't been edited yet, you can search for them from the Trove homepage search field. Enter with your keyword/s: <keyword> NOT has:corrections
Example: To find unedited articles that contain the word: jazz, enter in the Trove homepage search field: jazz NOT has:corrections
If you want to find items that can be edited, you can search for them from the Trove homepage search field. Enter with your keyword/s: <keyword> has:correctabletext
Example: To find the articles in the Australian Womens Weekly that you can edit, enter in the Trove homepage search field: Australian Women's Weekly has:correctabletext.
With great power comes a bit of responsibility. Please follow these guidelines to ensure that text in Trove is edited correctly wherever possible.
- Match your edits to the exact text of the onscreen article wherever possible.
- Don’t delete text from the transcript if it appears in the onscreen article.
- Include only the text of the onscreen article in your edits. If you have more information or thoughts to share, put them in a tag or note.
- You don’t need to match the visual style of the article with bold or italic font.
- If you find an edit that obviously doesn’t match the onscreen article, feel free to re-correct the text. If you prefer to advise Trove, get in touch.
Spelling and formatting
For spelling or formatting errors, add the corrections in a note that is separate to the transcript. Some spellings for personal or place names may reflect a historical spelling at the time that the article was written. Trove’s search engine also uses text in notes and tags, so an article with an added note will still be found in a search.
Example: If the name of Australian actor Lottie Lyell is consistently spelled "Lottie Lyall" in an article and you’re sure that it’s the same person, you should leave the transcript spelling as "Lottie Lyall" and add a note saying: "Lottie Lyall = Lottie Lyell", "Lottie Lyell actor", or similar. Trove’s search engine uses this note, so a search for "Lottie Lyell" will still find this article.
Sometimes it is not possible to match text exactly. For example:
- Transcribe upside-down letters as normal letters. Add a note letting others know that the line/letter is upside down when it is important – such as when the upside-down letters are the solution to a puzzle. If the upside-down letter is a printing error, no note is needed.
- The ‘long s’ or ‘ſ’ (as appears in many early publications) should be transcribed as a lowercase ‘s’.
- Transcribe names starting with M’ as Mc, or add a note with the corrected spelling.
- Minor errors, such as transposed letters in a word where the correct spelling appears elsewhere in the article, may be corrected.
There is no spell check in Trove's editing mode. Spell checks appearing here are from your web browser and you can change the options in your browser settings.
Punctuation and hyphens
Punctuation and hyphens should match the text as originally published, with two exceptions:
- Text in small caps (this looks like upper case, but in a smaller font) can be corrected to either all upper case or sentence case text.
- Hyphens that split words across two lines can be kept as is, or removed so the whole word is combined at the end of the first line. Neither way affects Trove’s search. Other kinds of hyphens should stay as is.
Sometimes it will not be possible to exactly match punctuation due to differences between historical printing equipment and modern keyboards. As punctuation does not affect Trove's search results, we advise choosing modern punctuation that matches as closely as possible and being consistent.
Sometimes, non-alphanumeric characters (e.g. @, # or &) appear in the transcript where a picture in an article has been incorrectly processed. If these characters do not appear in the article text, they can either be ignored or removed. They do not affect Trove's search results.
If you come across a character in the text that isn't available on your keyboard, select Insert symbol and choose from frequently used special characters. Characters not in this list can be copied from either the Character Map (Windows), or the Character Viewer (Mac).
Correct non-English text using the same guidelines as for English text. For Latin script languages (e.g. Estonian, French and German) most accented or special characters can be accessed through the Insert symbol button. These languages follow the standard horizontal left-to-right reading order.
Chinese language articles in Trove may have any of the following reading orders: horizontal left-to-right, vertical right-to-left or horizontal right-to-left. When editing, correct the text to the modern horizontal left-to-right reading order, as this will give the best results in Trove search.
If you would like to provide an English translation of an article, do this in a note and not in the transcript.
Tables may appear in the transcript as either: each line read left-to-right across the entire table or as columns one after another. Edit tables in the transcript to match how as they have been originally created in the article – if the table has been transcribed as left-to-right across the lines, separate the columns by using the vertical bar/pipe symbol: |
Note: This symbol can be found on a keyboard directly above the Enter key, although its position may vary on some keyboards.
You can only edit one page of articles at a time. If an article spans over 2 or more pages, it will have a separate transcript for the text on each of those pages. Save your changes on the current page before moving to another page.
Missing or combined text
The number of lines of text in a transcript should match the lines in the article. If a transcript is missing lines or has combined multiple lines into one, you can split or add lines. To do this, hover over the line you are editing until the + and - symbols appear below the right side of the line. Select + to add a new line below, or - to remove the current line.
Out of order text
If text in a digitised article appears to be out of order, this is often due to a digitisation processing error. Unfortunately, text order in an article can’t be changed after the page is loaded into Trove.
When editing a transcript for an article with out-of-order text, match the text to the article and don't use the transcript to correct the order. This is because each line of the transcript is linked to a specific location in the online article and changing the order will break these links. If a word or phrase isn’t searchable due to the order of the text, please add a note to capture this.
If the boundaries of an article have been incorrectly identified on the screen, and some of the text in the transcript belongs to another article, it should remain in that transcript so as not to break the links between the transcript and the article (as explained above). Once an article is loaded into Trove, it can’t be re-split or replaced. Add a note to explain which text belongs to which article.
Wrong page numbering
Page numbers in Trove are considered to be page sequence numbers, indicating the order that pages appear in a newspaper or gazette, rather than reflecting the actual page number. This may be due to an error in processing, where a second edition or supplement with separate page numbering has been included within the sequence of the newspaper, or the use of non-standard numbering for the newspaper eg 4a.
Poor quality or sideways pages
Articles in Trove are usually digitised using an original master microfilm. This means if the pages used to create that microfilm were of poor quality, damaged or were oriented the wrong way, then they will also appear that way in Trove. Currently, it’s not technically feasible to replace individual pages in Trove, so it is unlikely that these issues can be corrected, even if it is possible to find a better-quality copy of the original page.
Poor quality text often means poor quality transcripts. If you are unable to decipher the article text, either leave it as is, or use your best guess as to what the words are. Others may also be able to assist or may have information that allows them to better determine the correct word.
It may take up to 5 minutes for the Trove search database to update and for your edited text to appear in Trove searches.
The lines you have edited are updated in your Trove account if you are logged in while editing. If you make edits while you are not logged in to your Trove account, they cannot be added to your account later.
See and compare your edits
All edits you have made while logged into your Trove account can be seen in the Text Corrections section of your Trove profile. The default view shows all edits made, with the most recent editing session first, but you can change this through the Added within drop-down menu to only show edits from the last hour, week, month etc.
Each entry in the list is an editing session - beginning when you started editing and ending when you saved and exited the transcript. If you've come back to a transcript multiple times, you'll see multiple entries for the same article.
Under each editing session in your list, you will see Compare written in small blue text. Clicking on this text will open a list of what you've deleted and added from the transcript, highlighted in orange and green, respectively. This function is handy if you've started editing a long transcript and wanted to find your place so you can finish editing later. Clicking on title of the article will take you back to the article in the newspapers viewer.