Overly Aggressive Auto-correct

I’m writing a fantasy novel for NaNo this year, and I had planned to write it in Scrivener. But after using it to plan out some worldbuilding and character development, I’ve had to abandon it. I’ll have to use Word.

Why? I have characters with names like Tarin and Ila, and I’ve entered their names into the custom auto-correct. But whenever I type “tarin” or “ila” in any word whatsoever, the system capitalizes them (sTaring, dIlated). I can’t pound out my 90k-word manuscript and leave that chaos in there. It’s visually distracting and I will be in too much of a hurry to go back and change them manually.

Why doesn’t the auto-correct include an assumed space at the beginning and end of the words in its database? It’s baffling. I’ve adored Scrivener since I got the beta last November, but I’m frustrated and disappointed at this development. I simply can’t use it for typing rough drafts if it’s going to be this snotty about capitalization.

This sounds like something just needs to be tweaked in settings. I’m a little unclear where you’ve added the names–is this to the project’s Auto-Correction list (Project>Auto-Correction) or to your personal word list? I’m assuming the latter since the former should only be offering suggestions rather than doing the correction for you entirely, but even with the latter I’m not seeing what you’re getting. Could you post a screenshot of your Auto-Correction tab (from Tools>Options…) or list all the options that are checked? Also, can you check Help>About Scrivener and post which version number is listed at the bottom of that window?

You know what? I went looking for the auto-correct list and saw that there’s now a Personal Word List button. That wasn’t there when I added these names eons ago. I just put the names into the Substitution List.

I’m using version 0.0.45.

Here are my skishes:

Currently checked Options

Substitution list auto-capitalizing stuff

So…should I transfer all my names to the Project>Auto-complete list? Use the Personal Word List in the Options menu?

Setting up your substitution list will take care of this. I have a habitual mis-type that I’ve tried to rid myself of for years: I mean to type “the” and 8/10 of the time I type “hte.” Scrivener’s auto correct takes “hte” and turns it into “hate.” Drove me nuts until I edited the substitutions list to account for this AND made sure I have “allow additional substitutions” selected.



Ugh I am suffering from this thing as well. It won’t let me type the words I want but instead puts down the words it wants! :open_mouth: :open_mouth:

I had posted a similar question here: viewtopic.php?f=30&t=15198&start=0

This was part of my question:

Subsequently, I’ve turned off all of the auto-correct features because of how aggressive it was. But for me, what was less tolerable was how it seemed to make wrong, or undesireable, corrections as often or more frequently than it made correct corrections.

I have a few words of my own that I consistently type dyslexic-ly and other strange variations, as in the quote above. My question then was if the dictionary could be modified–as in deleting words–because adding new words wasn’t stopping it from making old, incorrect corrections (e.g., “dont” was corrected to “dint” rather than “don’t”; don’t is already correct and in the dictionary). In Word, words can be directly modified in the dictionary. I understood the answer to my question to be no because of how Scrivener’s dictionary is structured. Put in layman speak for me I was told it cannot function like Word’s dictionary for several reasons.

Russ, have you found a way to overcome that? It sounds like the autocorrect dictionary already had an incorrect correction for a word you’d type yet you’ve managed to tell it to autocorrect to something different? Or, put differently, when you’d type “hte”, it was giving you “hate”; however, now you got Scrivener to auto correct to “the”?

Sorry for the slow response, I haven’t had much time for forums these last few weeks.

I’m not sure I would go so far as calling it an “incorrect correction;” I’d hazard a guess that the person who assembled the dictionary list struggles with a typo where he/she types “hte” when the intent is to type “hate.” My particular idiosyncrasy is typing “hte” instead of “the.” It could be based on statistical occurrence - perhaps more people mistype “hate” than “the!” :astonished:

But yes, I have found a way to override the auto correct for specific instances by making use of the functionality of the substitution list, that will give you functionality very similar to Word’s auto-replace and custom dictionary scheme. Here’s the process:

Go to the main Options menu (the one that comes up when you push F12), and then to the Auto-Correction tab.
Click on the “Edit Substitutions” button.
Add an entry to the substitutions list, putting the typo you want auto-fixed in the “replace” column and the correct word in the “with” column.
Make sure the “Enable additional substitutions” option is checked.
Click OK, and go back to writing!

Now, whenever you commit that particular typo, the substitutions list will take precedence over the main dictionary. I can type “hte” as many times as I like and will always get “the” as the correction.

Did this answer your question?



Thanks Russ, I was able to figure it out shortly after posting the question.

On that same topic though, I have stumbled across another issue that I brought up (here, [url]Auto-correction substitution - curious behavior]) but have not seen any responses on yet.

I was using the auto-correction’s substitutions feature to catch non-standard names as well my habitual mistypings but have had to remove some from the list because of curious conflicts. Without any responses to my question, I am unable to determine if this behavior is considered normal or glitchy.

Thanks for the input Russ.

True, statistics will support one word being mistyped a particular way more than another but, from this writer’s perspective, the result of the auto-correction action is incorrect if it makes a selection that is not what I intended. Any number of substitutions may be correct for the computer to choose from a statistical perspective but when there is an actual word that someone is attempting to type (but mistypes) and the computer attempts to fix but assumes something different, the resulting word is considered incorrect. The actions I’m judging as correct or incorrect is whether the computer properly acted on the writer’s behalf.

Another author and I were trying to figure this out on Twitter the other day. Apparently, it is causing problems. Author problems. I’m considering just turning off spell check all together and moving things over to MS Word for that.