You’ve heard it said that “there are thousands of textual variants in the Bible. How can we know what the original text said?” But I say to you, “most of the variants are trivial, but anyway, use a good apparatus and follow the basic rules.”
What are the basic rules? How do we figure out which manuscripts of the Bible are changed by copyists and which manuscripts are the original? Before giving these rules, let’s look at a random example. And by random, I mean there are 789 pages in my Greek New Testament and I am going to use a random number generator to pick a page 1-789, then point at at the page with my finger.
p654, which is Titus 3:5-15. Let’s look at verse 7.
“so that being justified by His grace we would be ⸢made heirs according to the hope of eternal life.”
Notice the little ⸢ mark in front of the word “made.” There is a variant on this word. If you look down at the NA28 critical apparatus, it says this:
So here’s the problem: one set of manuscripts use γενωμεθα and the other uses γενηθωμεν. (Because the editors think the second one is better, they put it above in the body text, and here they just say txt). What’s the difference? They are both Aorist tense (past tense), both subjunctive mood (“would”), both 1st person plural (“we”). The only difference between these words is that γενωμεθα is in the Middle voice and γενηθωμεν is in the Passive voice. Since we have no Middle voice in English, there is literally no difference in our Bibles based on this variant.
That goes a long way to showing how trivial most of the variants are. And there are thousands more not listed by NA28 because they are deemed too trivial. The ones in the apparatus are the interesting ones, the ones that split the manuscripts, but even then, most are pedantic and do not change the passage’s interpretation.
But for sake of this example, I will explain the rules of how to decide based on the notes in the apparatus. Pretend like it makes a difference. Here is the apparatus code, deciphered:
א [two] is a correction of Sinaiticus from the 7th century.
D [two] is a correction of Claromontanus from the 9th century.
K Mosquensis from the 9th century.
L Angelicus from the 9th century.
Ψ Athous Lavrensis from the 9th or 10th century.
365. a manuscript from the 12th century. (Marked with an * because it is an especially reliable manuscript).
1241. a manuscript from the 12th century.
1505. a manuscript from the 12th century. (also marked with an *).
M. The entire rest of the manuscripts not listed.
א* [original] is the pre-corrected version of Sinaiticus, 4th century.
A is Alexandrinus from the 5th century.
C is Ephraemi Rescriptus from the 5th century.
D* is the pre-corrected version of Claromontanus, 6th century.
F is Augensis from the 9th century.
G is Boernerianus from the 9th century.
P is Porphyrianus from the 9th century.
33. a manuscript from the 9th century.
81. a manuscript from from 1044 (I wonder how they know this date)
104. a manuscript from 1087 (or this date) (Marked with an * because it is an especially reliable manuscript).
630. a manuscript from the 12th or 13th century. (also with an *).
1739. a manuscript from the 10th century.
1881. a manuscript from the 9th or 10th century. (with a *).
The first way to find the original is to use the dates of these manuscripts to see which is earlier. Since we have no manuscripts for the first spelling until the 600s, but four significant manuscripts from the years 300-500 for the second spelling, it would seem like the second spelling has stronger external evidence. External evidence is evidence from “outside” the text, i.e., from the manuscripts. So even though M, the rest of all manuscripts we have, agree with the first spelling, we go with the second. Manuscripts have to be weighed, not counted.
Internal evidence comes next. These are arguments from within the text. For example, scribes are more likely to add than to subtract from a manuscript. So if one is longer than the other, the short reading is preferable. Also, scribes have a tendency to skip words if two words end in the same letter. If they are looking from their left papyrus (the original) to the right papyrus (their project) and in the process they see του ενρου αβρου σμαλου τιπλου ασωμου μνενομου you can bet they will see the -ου endings and get mixed up in their spot on the page. Also sometimes scribes skip entire lines if two lines happen to start with the same letter. More importantly, scribes are more likely to add later doctrines into earlier texts than to delete proof-texts for their favorite doctrines (this is the story of 1 John 5:7). Finally, if one reading is more difficult than another, keep the difficult one. Scribes tend to simplify difficulties, not create new problems in the text. In this verse, I see no internal evidence to decide either way.
So, with moot internal evidence and strong external evidence for γενηθωμεν, we go with that. And so did the NA28, and so do modern English Bibles.
This example should show that the tons and tons of textual variants do not make it harder to know what the Bible originally said. The reverse: because there are so many, it is easier to recover the original text.