How to use Applescript text item delimiters to find and replace text in Pages?

I hope this will help you. This uses AppleScript's text item delimiters to split/join back texts. It can be more compact, but this is a comprehensive way to write it. As you can use it often in your script, it's a good thing to put it in a special subroutine. I build a list of pairs {search,replace} easier to maintain in one place, and a "repeat" loop to apply every pair of corrections. Don't forget the "my" statement as Pages doesn't own strRepl() and will fire an error.

Unfortunately, extracting text, and putting it back into Pages will loose any text attributes. So here it is :

set findReplaceList to {{"a", "A"}, {"b ", "B"}, {"this", "that"}}

tell application "Pages"
    set bodyText to body text of front document -- get the content as text
    repeat with thisFindReplaceValues in findReplaceList
        copy thisFindReplaceValues to {findItem, replaceItem} -- put first and second item resp. in findItem and replaceItem
        set bodyText to my strRepl(bodyText, findItem, replaceItem) -- search and replace text
    end repeat
    set body text of front document to bodyText -- put the new text back. Loosing attributes.
end tell
    
on strRepl(SourceStr, searchString, newString)
    set saveDelim to AppleScript's text item delimiters
    set AppleScript's text item delimiters to searchString -- change ATID : the search item
    set temporaryList to every text item in SourceStr -- split the text in parts removing searched items
    set AppleScript's text item delimiters to newString -- New ATID : the replace item
    set SourceStr to temporaryList as text -- this put back the parts to text with newString between
    set AppleScript's text item delimiters to saveDelim -- clean up ATIDs
    return SourceStr
end strRepl

This script is a variation on @Chino22's. Given the consistent requirement here (always a single letter, replaced by itself), I've moved to a simple list of single elements and set the replacement when calling the handler.

-- List of prepositions to seek out (added the 'z' as it was prevalent in the article used for testing)
set chList to {"v", "s", "k", "u", "a", "z"}

    tell application "Pages"
    set bodyText to body text of front document
    repeat with prep in chList
        -- call replacement handler
        set bodyText to my strRepl(bodyText, space & prep & space, space & prep & character id 160)
    end repeat

    set body text of front document to bodyText
end tell

on strRepl(srcStr, oldStr, newStr)
    
    set AppleScript's text item delimiters to oldStr
    considering case
        set temporaryList to every text item in srcStr
    end considering
    set AppleScript's text item delimiters to newStr
    set srcStr to temporaryList as text
    
    return srcStr
end strRepl

NB My search and replace strings include a space both before and after the letter. This ensures that only single-letter words are affected. I added a considering case to further restrict the search to lower case letters. The 'character id 160' specifies the non-breaking space. Finally I left out the first and last delimiter commands to reduce clutter. Add them back at your discretion. A single letter followed by punctuation will not be processed.

Regarding some of the errors you were seeing… They are likely a result of Pages having issues with text item delimiters within its tell block. In general, you would need to split the script into three sections, along these lines:

tell application "Pages" to set bt to body text of front document

myriad delimiters stuff, including 'set editedText to…'

tell application "Pages" to set body text of front document to editedText

Using the handler as Chino22 suggests circumvents this issue by putting all that work within the handler (which is outside the tell block). Also, 'beginning' and 'rest' don't mean what you assume they do in applescript. Finally, I have read of recommendations for working at the paragraph level rather than with the entire body text. It may not be an issue for you but perhaps if you are working with very large documents and have issues, it may be worth making some modifications to the script.