Dealing with UTF8 and loading data with SQL*Loader

I'm running Oracle 12.1.0.2 Standard Edition

We have files in UTF8 that we are trying to load in SQL*Loader. I've got 'CHARACTERSET' in the control file set to 'UTF8'. The database is created with AL32UTF8.

The file contains only standard English and some Spanish characters. The data is loading and characters are being interpreted correctly.

My problem is with the fact that the file contains fixed field length data, and the control file is using 'POSITION(x:y)' parameters, to define the fields. For lines that only contain standard, single-byte characters, everything works fine. However, the POSITION parameter uses byte semantics (regardless of the value of the 'LENGTH SEMANTICS' parameter). So, if a particular line has a two-byte character in it,every field following the field with the two-byte character is off by one. And, of course, if there are multiple two-byte characters in a particular line, that line will be off by n, where n is the number of two-byte characters in that particular line.

I'm no SQL*Loader expert.....am I missing something obvious here? Does anyone have any idea how to solve this?

Thanks,