I wanted to comment on the Qeri vs. Ketiv article, but can't seem to create an account - it thinks I want to log in, and of course doesn't recognize my as-yet-nonexistent name and password. So I'll post my comment here.
The article states - or rather implies strongly - that the q'ri marginalia are primarily a post-Masoretic phenomenon. While that may be the case for some instances, the Talmud actually discusses the phenomenon, several centuries beforehand (in the context fo the Talmud, the term "massorah" or "massoret" refers to the actual spelling of a word vs. its vowelization). The traditional Jewish approach, following that lead, is that the q'ri variances are a specific Rabbinic institution to "soften" the message and/or introduce additional midrashic meanings.
An example of "softening" can be found in Deuteronomy 28:27. The text reads, "The Lord will smite you with the boils of Egypt and with hemorrhoids..." where the word for that affliction is "'afolim," a word that literally refers to the anus. The q'ri renders it "t'horim," (with a guttural "h," the letter het), which describes the physiological phenomenon without overt reference to an "unclean" body part.
Just a few verses later (verse 30): "A man shall betroth a woman and another man shall lie with her..." where the verb "lie with her" uses the root "sh-g-l" in the text, which denotes an ongoing sexual relationship, whereas the q'ri has it read with the root sh-k-b(v), which refers to a one-time incident.
An example of the "broadened meaning" variety occurs in Psalms 100:3, the same "lo" variation that the article mentions in Job. In this case, the text reads, "Know that He is God; He made us, not we..." with the lamed-aleph spelling of "lo." The q'ri renders it with a vav, not an aleph, having it mean, "...He made us, and we are His..." What's interesting about this case is that there's no appreciable difference in the pronunciation between the two; it's clearly only referring to the "text."