In this guide

To count each character’s spoken words, gather all of that character’s dialogue into one passage, remove labels and directions, and count the remaining text. Repeat the process for every speaker, then put the results in a table so you can compare their contributions.
The most useful count starts with a few decisions: which version of the script you are using, what qualifies for your comparison, and how you will handle lines spoken together. Apply those decisions consistently, and keep a copy of the original script nearby so you can check anything that is unclear.
Set the scope and counting rules
Choose the script and the comparison
Start with one complete version of the script and note its title, version or revision date if available, and language. If you later compare your results with someone else’s, these details help you establish whether you counted the same text. When working with a translation, decide whether you want a count for that translation or for another language version; treat them as separate projects unless your purpose calls for combining them.
Next, decide what your table is meant to measure. A comparison of spoken dialogue might include only lines assigned to characters, while another project might also include narration or voiced inner thoughts. You can include these passages in a character’s total, give them separate rows, or leave them out. Write down your choice before collecting text, and apply it to every character.
Resolve ambiguous labels
Decide how to handle speech delivered by someone who is not visible, as well as dialogue assigned to a group. For a shared line, one option is to credit the full line to each named speaker; another is to list the line separately. The first option measures the words assigned to each person but counts the same passage more than once when character totals are added. A separate row makes it easier to count each passage once in the overall total.
Labels need context. In Chinese short-drama scripts, OS can denote inner voice, while O.S. in English screenplays usually means off-screen speech. Check the script’s own conventions before deciding how to classify a particular passage. If a label is unclear, mark it for review rather than silently assuming which category it belongs to.
Keep a short rule note beside your working table, such as: “English revision dated May 2; narration separate; shared lines in a group row.” Your note can be brief, but it should explain choices that affect the comparison.
Collect each character’s spoken text
Make a working passage for each speaker
Read through the script and create a separate working passage for each character. Copy that character’s included dialogue into the same passage wherever it appears, even if the lines occur in different scenes. Keep the original script intact so you can return to it when checking a line or a speaker.
Before you begin, make a list of the names and labels used for each person. If one character is called “May” in one scene and “M” in another, combine those lines only after you have established that both labels refer to the same character. Conversely, do not merge similar names without checking. A small alias list makes the collection process easier to audit.
Remove speaker labels, scene headings and stage directions from each working passage. Keep the dialogue itself, including words that form part of a spoken sentence. For example:
MAY: We should leave before the rain starts.
(May picks up her bag.)
JO: I will bring the blue umbrella.
MAY: Meet me by the door.
JO: Wait for me.
For a spoken-dialogue count, May’s passage would contain “We should leave before the rain starts” and “Meet me by the door.” Jo’s would contain “I will bring the blue umbrella” and “Wait for me.” The names and the parenthetical action are removed because they are not part of the dialogue being counted.
Keep uncertain material visible
When a direction appears inside a line, separate the direction from the words assigned to the speaker. If you cannot tell whether text is meant to be spoken, flag it and return to the surrounding scene. Record the decision you make so you can treat similar passages the same way.
For narration, inner voice or a shared line, use the category you chose at the start. If you are listing a passage separately, give it a clear working label instead of leaving it mixed into a character’s text. Once a passage is cleaned, check it against the script for missing lines or accidentally copied directions before moving on.
Count and record each passage
Count one cleaned passage at a time
Use a word counter with each character’s cleaned passage as the input. Read the resulting word count and record it against that character’s name. Repeat for every character and for any separate categories you decided to track. Use the same counter and approach throughout the comparison so that a change in method does not affect one character differently from another.
Before counting the full script, try a short passage whose words you can check yourself. Pay particular attention to contractions, hyphenated expressions and the language of the script. If the result does not match the convention you need, decide how you will handle those cases and make any adjustments consistently. Note your convention alongside the results; different counting approaches can produce different totals for the same passage.
For the short example above, a straightforward count gives May 12 words and Jo 9 words. The working table could look like this:
| Speaker or category | Word count |
|---|---|
| May | 12 |
| Jo | 9 |
| Total | 21 |
This example has no shared lines or separate narration. In your own table, add a row for each relevant speaker and any category you chose to keep separate. If you exclude a category altogether, note that choice outside the table so a reader does not mistake the total for every word in the script.
Preserve a useful trail
Keep the cleaned passages until you have finished checking the results. Label each one clearly, and record the counter and any special counting decisions with the table. If you revise a passage after finding a missed line, recount that passage and update its row. This small record helps you explain how you arrived at the numbers and repeat the process when the script changes.
Compare contributions and check the result
Calculate totals and shares
Add the character counts to get a total for the passages represented by those rows. Include separate narration or group rows in a broader total only if that matches the scope you chose, and label the result accordingly. If a shared line was credited in full to multiple characters, the sum of their counts includes that line multiple times. Explain this when reporting the total, or keep shared speech separate if you want each passage counted once.
To find a character’s share, divide that character’s count by the relevant total and multiply by 100. In the example, May’s share is 12 ÷ 21 × 100, or about 57.1%; Jo’s is 9 ÷ 21 × 100, or about 42.9%. Both shares use the same total of 21 spoken words. If you round percentages, their displayed sum may differ slightly from 100%.
Choose the denominator to match the question. If you want to compare characters’ dialogue with one another, use the character-dialogue total defined by your rules. If you include narration or a separate group row in the denominator, say so: the resulting shares answer a broader question about the included passages.
Check the collection before sharing
Go back through the script and confirm that every passage within your chosen scope appears in the right working text or separate category. Look especially for characters introduced under another name, short responses, and lines that continue after an action direction. Check that excluded material was excluded consistently and that speaker labels or scene headings did not slip into a passage.
Then verify the arithmetic: do the individual rows add to the stated total, and do the shares use that same total? If you credited shared lines to multiple speakers, check that the explanation accompanies the figures. Revisit any flagged passage and document its resolution.
A word count is a useful measure of how much text each character is assigned under your rules. The table becomes much easier to interpret when it travels with the script version, the treatment of special passages and the counting convention used.