Post History

There are actually multiple questions here, each with a different (though related) answer:

Q1. **Why does the last line disappear and/or the character count shrink _In the example you provided in your screenshot_?**

Answer: **It doesn't.** Look at the image you posted. The number of lines is identical (and is four, not eight _or_ nine), as is the length of the last line. The only difference between the number of characters (or even the number of times each different character appears) is (presumably due to a typo and/or proofing error) an _additional_ (and unpaired, so structurally incorrect) right-quotation mark in the second (supposedly shorter) version.

Q2. **Okay, so why does the author say that the length and character count change? Is she just crazy?**

Almost certainly not. What has pretty clearly happened here is that:

1. The author typed up her original examples in (some specific version of) Word (with some specific settings). Her claim was accurate with respect to that initial version.

2. The published version with typeset (a) in some other software, (b) in a later version of Word, perhaps with slight tweaks to the relevant algorithms, and/or (c) with a different set of options selected. This may have happened in the very first version, or been introduced later.

3. Rather appallingly, given that tricks for squeezing information into fewer pages while adhering to courts' filing requirements (many of which are highly sensitive to the exact details of the software used) is _the entire point of the book_, **no one checked** to make sure that the statements in the body text and captions still applied to the examples in the tables after this change in typesetting.

Q3. **Okay, so why did changing the order of those phrases change the length of the paragraph in the authors' original version?**

Text has to fit on lines of a fixed width, and modern practice does not allow simply placing characters until you reach the end of the line, and then continuing on the next line, regardless of whether or not you're in the middle of a word. Consequently, part of the duty of a typesetter (whether human or (far more often these days) mechanical) is to fit as much of the text onto each line as they can while producing an acceptable result.

In its simplest form (left-justification with no "clever" options), this can mean adding whole words (and their attached punctuation) to a line until the next whole word would be too long to fit on that line, and the moving down and placing that whole word on the next line. In its more complicated forms (full-justification, (auto-)hyphenation, widow and orphan control, flexible margins, etc.), the details are more elaborate, but the basic principle remains: how much empty space ends up in the lines of a paragraph depends not only on what words are in that paragraph, but on what order those words come in, because a short word can fit onto the end of a line where a longer word wouldn't.

Q4. **Okay, but why did the character count (originally) change?**

The author most likely had auto-hyphenation on, meaning that a word that would otherwise be forced onto the next line can instead be split by a hyphen, with the first part + hyphen on the first line, and the remainder on the next line. The "A. Original Text" most likely had three more auto-inserted hyphens than the "B. Revised Text."

On a tangential note, auto-hyphenation (like full justification, which allows inter- and intra-word spacing to vary (up to a certain point) in order to achieve lines that are all the same width (except for the last line of a paragraph)), tends to _reduce_ the difference in length that results for doing things like switching word order, but it doesn't (necessarily) eliminate that difference entirely. That's partly because hyphens themselves take up some space, and partly because there are rules for where hyphens can be inserted for this purpose (only between syllables, not in a word that already has a hyphen as part of its standard orthography, etc.), and additional guidelines that strongly disfavor inserting them in some places that they _could_ be inserted, such that well-designed software will not insert a hyphen in such a place unless the alternative would be extremely bad (e.g., pushing a single short word onto a new page that would otherwise have no text on it at all).