In Job 40, the Lord is infallibly describing a real historical creature, called ‘Behemoth’. No known living animal, such as the elephant or hippopotamus, fits the passage adequately. A detailed analysis of the key clause Job 40:17a suggests that the most natural interpretation is that the tail of Behemoth is compared to a cedar for its great size. Consequently, the most reasonable interpretation is that Behemoth was a large animal, now extinct, which had a large tail. Thus some type of extinct dinosaur should still be considered a perfectly reasonable possibility according to our present state of knowledge.
The passage Job 40:15–24 gives a detailed description of the creature named ‘Behemoth’. There has been controversy as to what Behemoth really is. This paper focuses on the clause Job 40:17a, which is crucial to the proposition that Behemoth could have been a type of dinosaur. The other common proposals are also analyzed, and some general considerations are made concerning the whole passage.
The word ‘Behemoth’ (Job 40:15) is literally a plural form of a common Old Testament (OT) word meaning ‘beast’. However, practically all commentators and translators have agreed that here we have an intensive or majestic plural, so that the meaning is something like ‘colossal beast’. This case is similar to the word ‘Elohim’ (the most common name of God in the OT), which is actually a majestic plural form, but is always used with a singular verbal form, just as is encountered in this passage. Also, we read in verse 19 that Behemoth was the ‘chief of the ways of God’, which suggests that Behemoth was one of the largest (if not the largest) of God’s creatures.
There are basically three animals that have been put forward as candidates for Behemoth which we will consider in this paper: the elephant, the hippopotamus, and some type of dinosaur.
Gordis gives an interesting overview of how Behemoth has been interpreted throughout history, noting that ‘the interpretation has oscillated through the centuries between two poles, mythical and real’.1 The interpretation that Behemoth was an elephant is very old and was common among mediaeval scholars. The original KJV-1611 had a marginal note ‘Or, the Elephant, as some thinke [sic]’. But the French Protestant pastor Samuel Bochart in his Hierozoicon (1663), which analyzed the animals in the Bible in great detail, suggested that Behemoth was a hippopotamus, and this has since remained as the most common interpretation. The suggestion of a dinosaur has only arisen in relatively modern times.
We will first dismiss the idea that Behemoth is some kind of mythological creature. Gordis gives many good reasons why Behemoth cannot be interpreted in this way. For example:
‘the First Speech of the Lord deals with flesh-and-blood animals and birds, from the existence of which important conclusions are drawn regarding the nature of the world and man’s place in it. … The same consideration supports the idea that Behemoth and Leviathan are also natural creatures, the existence of which heightens the impact of God’s argument.’2
The descriptions of the physical and behavioural characteristics of Behemoth are also not consistent with mythological creatures.
At this point we will also dismiss the idea, common amongst 19th century commentators (e.g. Delitzsch3 and Gesenius4), that the word ‘Behemoth’ actually comes from an Egyptian word p-ehe-mou meaning ‘ox of the water’, and thus refers to the hippopotamus. But Driver and Gray reply: ‘there is no evidence that the often cited p-ehe-mou ever existed’.5 Likewise, Gordis rejects the proposed Egyptian word; he states that it just doesn’t exist!6
Now if we believe in the inspiration and inerrancy of the Bible we must maintain that the words spoken about Behemoth (and also Leviathan), were spoken by their Creator, who would have known the intimate details of His own design. The description of the animals in chapters 38 to 41 is given to convince Job of his ignorance and folly. It is thus a critical observation that, when the purpose is to show how marvellous an animal is, surely the most amazing facts about that animal, and the ways in which it is different in habits or appearance from all others, should be stated.
The key clause
Bearing all this in mind, we will now consider the key clause in Job 40:17a. First of all, here are some of the English translations of the clause that have been done:
- KJV: He moveth his tail like a cedar.
- NASB: He bends his tail like a cedar.
- NIV: His tail sways like a cedar.
- NKJV: He moves his tail like a cedar.
- REV: His tail is rigid as a cedar.
- RSV: He makes his tail stiff like a cedar.
- RV: He moveth his tail like a cedar.
Older translations are also of interest:
- Septuagint (Greek): He sets up/erects (estesen) his tail like a cypress.
- Vulgate (Latin): He ties up/binds (constringit) his tail like a cedar.
- Luther (German, 16th C): His tail stretches (streckt sich) like a cedar.
- Statenvertaling (Dutch, 17th C): According to his pleasure (Als ‘t hem lust), his tail is like a cedar.
- Diodati (Italian, 16th C): He raises (rizza) his tail like a cedar.
As can be seen, the translations have varied significantly! The reason is that the Hebrew verb in this clause is very problematic, so we shall study it in some detail.
The Hebrew verb in this clause is yachpôts. First of all, note that there is certainly a common OT verb châphêts, meaning ‘delight’, and yachpôts could be the Imperfect Qal form of this root. However, according to most lexicons and commentators, yachpôts here is actually a form of a different root, which occurs only here in the OT! The celebrated Brown, Driver and Briggs Hebrew lexicon (BDB) presents the two roots as follows:7
- A) châphêts: means ‘delight, take pleasure’, usually with preposition be (‘in’), thus: ‘delight in’. A proposed Arabic cognate word means ‘be mindful, attentive’.
- B) châphâts: means ‘bend down’. A proposed Arabic cognate word means ‘lower, depress’. They then translate our clause: ‘he bendeth down (extendeth down stiffly) his tail’.
The much more recent and widely acclaimed lexicon of Koehler and Baumgartner, which takes advantage of the advances that have been made in Semitic linguistics since the BDB lexicon, still presents a similar assessment, with two different roots:8
- A) châphêts: means ‘desire’. A Syrian cognate means ‘to try to get’, and an Arabic cognate means ‘to keep, to take care’.
- B) châphâts: has an Arabic cognate meaning ‘to make lower’. They then suggest that the verb in our verse, coupled with the word zânâb (‘tail’) means ‘to hang’, while an alternative translation is ‘to hold stiff’.
Most commentators basically agree that there are two separate roots. James Barr notes that the philologist Felix Perles suggested in 1895 that the verb here is not root (A) but root (B) with meaning ‘make straight, stretch out’.9 Driver and Gray state that root (A) is a different root from root (B) and give the same Arabic cognate as BDB for root (B), with meaning ‘lower, depress’, which is used in the Koran in the context of a bird lowering its wings (‘here, it seems, somewhat more generally, to bend’).10 Jenni and Westermann suggest that our verse has a different root from the common ‘delight’ root (A), and suggest the sense ‘to let hang (?)’ for root (B).11
Now most commentators also assume at the outset that Behemoth is a hippopotamus.Now most commentators also assume at the outset that Behemoth is a hippopotamus. Some of these then admit that our verse does not fit this very well. Andersen states that the hippopotamus fits the passage best, but then says: ‘But it is hard to see how his tail can be compared to a cedar’.12 Pope says: ‘As applied to the tail of hippo the hyperbole is extreme since the tail is absurdly small …’.13 Most absurdly of all, Hanson and Hanson say: ‘Our author must have been misinformed about the hippopotamus’ tail’!14
In contrast, other commentators, still labouring under the hippopotamus assumption, try to make the meaning of the verb fit this assumption! Many give to verb (B) the sense of ‘harden’ or ‘stiffen’, so that the whole verse is interpreted as meaning that Behemoth’s tail is compared in some way to a cedar tree for its stiffness or hardness. This idea is followed by Rowley,15 BDB (translation for (B), above), and Koehler and Baumgartner (alternative translation for (B), above), for example, and is probably the idea behind the translations of the REV, RSV and Vulgate versions of the Bible, quoted above. But there are no cognates given in Arabic or other languages to justify the sense of ‘harden’ or ‘stiffen’. It seems that this translation has only been proposed by those who make the a priori assumption that Behemoth must have been a hippopotamus! Furthermore, it doesn’t fit anyway, since the hippopotamus’ tail is short and fat (as Andersen, etc. above realized)!
Others have suggested the idea ‘stretch out’ (e.g. Perles, above). Gordis states that the verb ‘is generally derived from the Arabic hafasa “lower, depress”, hence “bend, arch” (P.), but the reference to the cedar makes this meaning inappropriate’.6 But he then dogmatically states that the context (i.e. his assumption that Behemoth is the hippopotamus!) requires the meaning ‘stretch out’.
So what can we make of all this?
The verb châphêts (A) (‘delight’) normally takes the preposition be (‘in’, ‘with’), but that is missing in this clause, which does give weight to the view that we do not have root (A) here. Also, because the preposition kemô (‘like) is normally used adverbially, so is tied to the verb, the ‘delight’ sense is dubious: it is hard to know how the animal could ‘delight in its tail like a cedar’ (with the ‘like a cedar’ referring to the manner in which it delights!). Note that the Dutch Statenvertaling is the only translation quoted above which interprets the verb in this way, but the whole clause is rather meaningless in this translation.
The sense ‘hang’ proposed for root (B) (e.g. by Koehler and Baumgartner, above) at least has a sense related in some way to its proposed Arabic cognate meaning ‘to make lower’. Furthermore, all translations and commentators quoted above have tended to assume that the creature Behemoth is the subject of the verb and ‘his tail’ is the object of the (presumed transitive) verb. However, if the ‘hang’ sense were admissible, then ‘his tail’ could be the subject of the verb, so that the clause could be translated ‘His tail hangs like a cedar’. This would certainly be valid grammatically, if the verbal sense were correct, but it remains difficult to give a good sense to the clause if the expression ‘like a cedar’ must be tied adverbially to the verb.
We must also bear in mind that the vocabulary of Job has several distinctive features. Pope notes that Job has more hapax legomena (words which occur only once in the whole Bible) than any other Biblical book.16 Driver and Gray note that Job contains several Aramaic words,17 and Young also stresses the ‘significant occurrence of Aramaisms’ and concludes: ‘In any case it is evidently true to say that a free expression of dialectal features is characteristic of the poetic style of Job’.18
In conclusion, it seems very unlikely that we have the ‘delight’ root (A) here, while we cannot be sure what root (B) precisely means, since the only proposals are based on tenuous connections with other Semitic cognates, and there is also a good possibility that root (B) is some dialectal word not in standard Biblical Hebrew. But from current philological considerations alone (ignoring the context of the whole clause), we can say that the only meanings of the verb which might be admissible are the ‘delight’ concept or something like the ‘lower’ or ‘hang’ concept. There is absolutely no philological justification for the ‘hardness’ or ‘stiffness’ idea!
Finally, note that the Septuagint (LXX), as quoted above, translates the key verb as ‘erects’, and it is just possible that the LXX translators were following an interpretative tradition which was based on a better knowledge of the Hebrew word or its origin, so that this would be the true sense of the word. But this is pure speculation—the LXX translators most probably just guessed at the meaning of the verb here, as most other translators appear to have done.
The tail and the cedar
Moving on from the verb now, how could one compare a tail to a cedar?
Since several passages in the OT make a comparison with cedars because of their great height (e.g. 2 Kings 19:23, Isaiah 2:13, Isaiah 37:24 and Ezekiel 17:22, Ezekiel 31:3), it may be that in Job 40:17a the tail is compared to the cedar for height. This could suggest the interpretation that Behemoth’s tail is lifted up high. We noted previously that the LXX sense of ‘erects’ could just possibly be the correct sense. It would give a reasonable sense to the whole clause, as animals can lift up their tails, and this might be applicable to some kinds of dinosaurs. But it must be emphasized again that this is pure speculation.
Another thought is the following: since the book of Job is independent in several ways from the other OT books (that is, it does not mention Israel, etc.) and is probably old, the literary expressions used in it need not be strictly interpreted in the light of the common imagery found elsewhere in the OT. Thus the vertical height of the cedar itself need not be stressed here. So an alternative interpretation which is quite natural is that the cedar is mentioned simply for its great size or length, and the fact that it is vertically high is not necessarily significant, as a cedar lying horizontally on the ground is just as much a cedar as one standing vertically. In Psalm 92:12, we read that the righteous shall flourish like the palm tree and ‘shall grow like a cedar in Lebanon’. Here the key idea is simply that of great size and strength, and height is not important.
At this point we should also note that some have suggested that the word ‘tail’ here could be interpreted as ‘trunk’ (see, for example, the footnote for this verse in the NIV19), so that Behemoth would thus be the elephant. But the most glaring objection to this idea is that we read later in verse 24 about Behemoth’s nose (and that it cannot be pierced)! The animal cannot have a trunk (described by ‘tail’) and also a distinct nose! Since the Lord is infallibly speaking in this passage, why would He not have used ‘nose’ in verse 17 if Behemoth is the elephant? Also, the word for ‘tail’ here (zânâb) occurs eight other times in the OT,20 and in each case the key idea is the base or end of something, or inferiority (which is also the natural way one uses the word figuratively in English). So it would be extremely unnatural in itself to interpret the word ‘tail’ as a ‘trunk’, whatever the context. The elephant’s tail proper is as inappropriate as the hippopotamus’s tail for the comparison with a cedar.
We saw above that there is absolutely no etymological or philological justification for the sense ‘harden’ or ‘stiffen’, while something like the sense of ‘hang’ might be possible, and the LXX also suggests the sense of ‘erects’. So it seems that the only reasonable way that one could compare the tail of an animal to a cedar as a distinctive feature of the animal is either for its height (assuming it is lifted up) or simply for its great length or size. The key point is that some significant comparison is made with the cedar (whatever the precise meaning of the verb may be) in order to highlight a distinctive feature of Behemoth.
The elephant’s and hippopotamus’s tails are thus completely inappropriate for such a comparison. But in contrast, the idea of a dinosaur with a very large tail is quite appropriate. In particular, specimens of sauropods have been found which have very large tails worthy of comparison with the cedar tree.
The whole passage
Finally, we will very briefly examine the suitability of the proposed animals in the light of the whole passage.
The elephant is outstanding for its trunk, its great size (especially its feet), its enormous appetite and its ears. None of these unique features are mentioned in our passage, but they ought to have been, if Behemoth was the elephant. Also, the elephant retreats to the depths of the forest during the hot part of the day.21 This does not seem to fit with Job 40:21, which suggests that Behemoth spends his time in marshy areas.
The hippopotamus is noted for its weight, its large and strong mouth, with its deadly tusks, its thick skin, its pink sweat and its ability to walk on the bed of a river for long periods. It spends most of the day in the water, as its skin dries out very quickly in the sun.22 Again, none of these unique features are mentioned in our passage! Similarly, the hippopotamus stays in the deeper water, and this does not seem to fit with Job 40:22, where we are told that Behemoth stays under the trees on the bank.
The main features of the dinosaurs are unknown, apart from the size of their bones, which indicates that some of them were much larger than any known land animal alive today. Consequently, because of our ignorance here, there is nothing in the passage to eliminate this possibility!
It is not surprising that before fossils of large extinct animals were found in great numbers, older conservative commentators only tried to identify Behemoth with some of the largest known living animals (even though none of these animals are suitable). The possibility of very large extinct animals did not really occur to them!
The whole passage in Job 40 concerning Behemoth certainly suggests a large animal, and no known living animal fits the passage adequately (for various reasons, including the detailed habitat presented).
The most natural interpretation of the key clause Job 40:17a is that the tail of Behemoth is compared to a cedar for its great size, and there is nothing in the context which contradicts this possibility, even though the exact sense of the verb is extremely difficult to determine.
Consequently, the most reasonable interpretation (which also takes the whole passage into account) is that Behemoth was a large animal, now extinct, which had a large tail. Thus some type of extinct dinosaur should still be considered a perfectly reasonable possibility according to our present state of knowledge.
I would like to thank Noel Weeks for reading a draft version of this paper and for several helpful suggestions.