A few lines in Troilus and Criseyde remind me of a Hebrew Bible verse. I want to compare the language of the book to the language of the verse, and I'm assuming he wasn't reading it in the original Hebrew.
I added the caveat “almost certainly” because stories from the Bible were popular subjects for retellings, so that in any particular case where Chaucer refers to a biblical passage, there are other sources from which he might have got the story.
For example, the Summoner’s Tale alludes to Matthew 23:7–8:
‘No maister, sire,’ quod he, ‘but servitour,
Thogh I have had in scole swich honour.
God lyketh nat that “Raby” men us calle,
Neither in market ne in your large halle.’
Chaucer might have got this directly from the Vulgate text of Matthew, but he could also have got it second-hand, for example, from the Roman de la Rose:
Et ament que l’en les salue
Quant il trespassent par la rue,
Et veulent estre apelé mestre,
Ce qu’il ne devroient pas estre
Jean de Meung (c. 1280). Le Roman de la Rose, volume 2, lines 12570–12573. Paris: Firmin Didot (1864).
Grace W. Landrum considered all such allusions in order to determine whether Chaucer’s wording is closest to the Vulgate or to other sources, and concluded that
After the subtraction of second-hand borrowings there remain unquestionably to Chaucer’s credit about 275 cases of direct dependence upon the Vulgate.
Grace W. Landrum (1924). ‘Chaucer’s Use of the Vulgate’. In Proceedings of the Modern Language Association 39:1, p. 99.
For example, in the passage from the Summoner’s Tale there are a couple of wording choices that support Chaucer’s source being the Vulgate and not the Roman de la Rose:
|Summoner’s Tale||Vulgate||Roman de la Rose|
|in market||in foro [in the marketplace]||par la rue [through the street]|
No single case is conclusive (since there could be another source not considered) but the weight of the evidence is strongly in favour of Chaucer having access to, or maybe even owning, a copy of the Vulgate.
Other versions of the Bible are extremely unlikely as sources for Chaucer: there is no indication that he knew Greek or Hebrew, and knowledge of these languages was poor among Western Europeans until after the emigration of Byzantine scholars following the fall of Constantinople in 1453.
Some people have suggested that Chaucer might have read the earliest (1382) version of Wycliffe’s English translation of the Bible. Walter Skeat, for example, suggested that we compare the Parson’s Tale with the Wycliffite translation of Genesis 3:1–7.
|Parson’s Tale||Early Wycliffe||Vulgate|
|how that the serpent, that was most wyly of alle othere bestes that god hadde maked, seyde to the womman: ‘why comaunded god to yow, ye sholde nat eten of every tree in paradys?’||But and the edder was feller than ony lifers of the erthe, the which made the Lord God. The which seide to the woman, Whi comaundide God to ʒow, that ʒe shulden not ete of ech tree of paradis?||Sed et serpens erat callidior cunctis animantibus terræ quæ fecerat Dominus Deus. Qui dixit ad mulierem: Cur præcepit vobis Deus ut non comederetis de omni ligno paradisi?|
|The womman answerde: ‘of the fruit,’ quod she, ‘of the trees in paradys we feden us;||To whom answeryde the woman, Of the fruyt of trees that ben in paradis we eten;||Cui respondit mulier: De fructu lignorum, quæ sunt in paradiso, vescimur:|
|but soothly, of the fruit of the tree that is in the middel of paradys, god forbad us for to ete, ne nat touchen it, lest per-aventure we should dyen.’||of the fruyt forsothe of the tree that is in the mydil of paradis, commaundide us God, that we shulden not eten, and that we shulden not towche it, lest perauenture we dien.||de fructu vero ligni quod est in medio paradisi, præcepit nobis Deus ne comederemus, et ne tangeremus illud, ne forte moriamur.|
|The serpent seyde to the womman: ‘nay, nay, ye shul nat dyen of deeth;||Forsothe the eddre seide to the woman, Thurʒ deth ʒe shal not die;||Dixit autem serpens ad mulierem: Nequaquam morte moriemini.|
|for sothe, god woot, that what day that ye eten ther-of, youre eyen shul opene, and ye shul been as goddes, knowinge good and harm.’||God forsothe wote, that in what euer day ʒe eten therof, ʒoure eiʒen shul be openyd, and ʒe shal ben as Goddis, knowynge good and yuel.||Scit enim Deus quod in quocumque die comederitis ex eo, aperientur oculi vestri, et eritis sicut dii, scientes bonum et malum.|
|The womman thanne saugh that the tree was good to feding, and fair to the eyen, and delytable to the sighte; she tok of the fruit of the tree, and eet it, and yaf to hir housbonde, and he eet;||Thanne the woman saiʒ that the tree were good, and swete for to ete, and fayre to the eijen and delitable in the siʒt; and she toke of the fruyt of it, and ete, and ʒaue to hire man, the which ete.||Vidit igitur mulier quod bonum esset lignum ad vescendum, et pulchrum oculis, aspectuque delectabile: et tulit de fructu illius, et comedit: deditque viro suo, qui comedit.|
|and anoon the eyen of hem bothe openeden. And whan that they knewe that they were naked, they sowed of fige-leves a manere of breches to hiden hir membres.||And the eiʒen of both ben openyd; and whanne thei knewen hem silf to be nakid, thei soweden to gidre leeues of a fige tree, and maden hem brechis.||Et aperti sunt oculi amborum; cumque cognovissent se esse nudos, consuerunt folia ficus, et fecerunt sibi perizomata.|
There are substantial similarities between the two English versions of the temptation story, but there are also a few places that suggest Chaucer was working directly from the Vulgate (or some other source deriving from the Vulgate) and not from Wycliffe:
|Parson’s Tale||Early Wycliffe||Vulgate|
|serpent||edder||serpens [snake, creeping animal]|
|alle||ony||cunctis [all, whole]|
|every||ech||omni [every, all]|
|nay, nay … nat||not||nequaquam [not at all]|
|harm||yuel||malum [evil, harm]|
|housbonde||man||viro [man, husband]|
Landrum was skeptical of the idea that Chaucer had relied on Wycliffe:
Furthermore, comparison of Chaucer and the Wycliffite versions in practically all other cases brings out great unlikeness of phrasing and diction.
Landum, p. 87.
Gareth Rees's answer has given a very nice, detailed summary of analysis by Landrum (1924) and also mentions Skeat, who was earlier. I just want to point out in this answer that there may be another possibility that Landrum probably could not have known about at the time, which is that Chaucer could have been working from a Diatessaronic source. The Diatessaron was a gospel harmony, originally in Syriac, that was also the source of gospels in many other languages. In various times and places, it was more common than the Vulgate -- certainly in Syria, but also probably in other places. We don't know for sure because it was later systematically suppressed. There is a Middle English gospel called the Pepysian Harmony. It's believed to date from ca. 1400, and to have been translated from an Old French Diatessaron that is now lost. This puts it at right around the correct time period to have been used by Chaucer. The text was only published in an edition by Goates in 1922, so Landrum, whose paper was published in 1924, may not have known that it was a possibility. This is probably a long shot, but it might be a fun project to go through Landrum's analysis and see whether anything can be said about any Diatessaronic sources. For example, Chaucer could have had access to both the Vulgate and the Pepysian Harmony, in which case some of his text might echo one and some the other. Googling on stuff like "pepysian harmony chaucer" didn't seem to turn up any previous studies of this possibility.
Chaucer was an educated man who would certainly have read the Bible in Latin. Interestingly, the date of Wycliffe's translation into English of parts of the New Testament more or less coincides with Chaucer's completion of Troilus and Criseyde in the mid 1380s. It might be worth checking for a correspondence there.