Posts Tagged ‘semantics’

Blood 1So what does hilas- actually mean? Morris didn’t have the linguistic chops to do the word-study. What would it look like if he had?

We are going to employ some of the basic tools of lexical semantics, nothing too technical, and see what we can learn about this term. (For a brief Dummies Guide to the tools we’ll be using, see here.)

Getting set up

First, we need to keep in mind that we are doing a word-study not a theological study. So at the end we are not going to be ready to answer the question in the title. Sorry. But hopefully we will be able to say what hilas- means in the NT.

Second, we are studying a Greek term, not an English one: hilas- not ‘propitiation’. So we don’t know what it means yet. And there’s no one we can ask. The only authority on NT Greek alive today is Jesus of Nazareth, and he is not currently taking questions. There is no authoritative book of definitions we can fall back on. We just have to figure out what it means from the way it’s used. Just like the guys who write the lexicons. So it’s all about usage, folks.

Third, there is probably more than one meaning for this term. There usually is. This is called a word’s semantic range. And when we look at the hilas- word group, there certainly seems to be a range of meanings. So we’re not going to end up with a tidy one-word answer that will make a nice chapter-title. Again, sorry.

sacred texts

sacred texts

Fourthly, we need to choose our territory. We are particularly interested in what hilas- means in the New Testament. But Morris was right to start his study in the LXX – because this is the literary context in which the NT documents were written. These are their sacred texts: they are massively influential. Simply put, the apostles’ writings are saturated with the language and imagery of the LXX. The NT documents are Jewish writings through and through. If you think in terms of concentric circles of context, the LXX is a close-in circle of context. Other domains, such as pagan/classical usage, are much further out and less influential. So we are going straight to the LXX to find our meanings there.

In the Old Testament LXX

Scholars agree that in the LXX hilas- very often conveys the sense of ‘cleansing’ or ‘expiation’. Here’s why: because it often has ‘sin’ as its object. In the usage, generally hilas- acts upon sin to make it better in some way. In other words, we’re talking cleansing, forgiveness, purification, purging – something of that sort is demanded by the usage.

However in some contexts a propitiatory sense seems to be in view, and at times perhaps both meanings are present. We will return to this.


a network of buddy-words

Nearly half of all the occurrences of hilas- in the LXX version of the OT are concentrated in two books: Leviticus and Numbers (76 out of 166 occurrences). The most frequent form of the word in the OT is exilaskomai (96 occurrences). So we will start with this term as it is used in Leviticus and Numbers.

In this context exilaskomai hangs out with buddy words like ‘sin’, ‘offering’, ‘release of sins’, ‘priest’, ‘blood’, ‘slaughter’, ‘altar’, ‘clean’, ‘purify’. These words create a field of meaning: they become a context for each other that influences the meaning of each word in the field. In these books, exilaskomai has a specific Levitical meaning, a strongly sacrificial flavour. It is practically a technical term for priestly operations.

One piece of grammar is very revealing here: in this priestly setting exilaskomai never has a personal direct object: i.e. there is no offended party in view being acted on. Rather it is the sins of the people which are acted upon: exilaskomai is what happens to their guilt. The categories at work are clean and unclean, not wrath and peace. Here exilaskomai means ‘cleansing from sin’ or ‘expiation’.

The other main forms of hilas- that occur in this Levitical context also relate to sacrifice. The hilasterion is a part of the ark: the ‘mercy seat’ or ‘atonement cover’ where blood offerings are sprinkled once a year on the day of atonement. And that day itself is called the day of hilasmos: ’the day of cleansing/atonement.’ This was a special day for acknowledgement of sin and sacrifices of cleansing. On this day offerings taken into the hilasterion achieve exilaskomai. So the three terms are closely linked, centred around the Levitical system of sacrifices and especially the day of atonement.

So we can say that in these sacrificial contexts in the LXX, hilas- nearly always refers to cleansing not propitiation: the removal of sin, not the turning away of wrath. This reflects the role the Levitical sacrifices have in the life of Israel. Leviticus/Numbers makes it very clear that the sacrificial system is established for the cleansing of unintentional sins:

When anyone sins unintentionally in any of the LORD’S commandments about things not to be done, and does any one of them: … he shall offer for the sin that he has committed            (Leviticus 4:1-3 – see also the comment on this in Hebrews 9:7).

In the Torah, these sins in ignorance are never said to arouse God’s wrath. God’s wrath is only ever aroused through Israel’s deliberate unfaithfulness. But dealing with rebellion like this goes well beyond the paygrade of the sons of Aaron. Sacrifice is not adequate to propitiate an angry God. It was never intended for this purpose.

Elsewhere in the OT, on occasions when God’s wrath is aroused, hilas- terms can be used to mean ‘propitiation’. Compare the very different use of hilaskomai in Exodus 32:14. Moses is on the mountain pleading for the people: a non-sacrificial context. The people have deliberately sinned with the golden calf, and Yahweh plans to destroy them. But then we read:

And Yahweh was hilasthe concerning the disaster that he planned to bring on his people.

Note the differences here: there is a direct personal object: Yahweh. The verb affects him: God is turned away from the evil he had planned for Israel. Here it is not cleansing which is in view: it is all about God’s anger. This purely propitiatory usage is fairly rare.


circles of context

We have seen two outer limits of the usage of hilas-: simple cleansing and simple propitiation. In most other occurrences of this wordgroup in the OT, things are not so simple. Sin, not God, is usually the object, the thing acted on. This suggests a widespread expiatory meaning. However it is often possible to argue (as Morris does) that ideas of wrath are hovering about in the wider context, flavouring hilas- with a ‘background’ propitiatory sense.

How strong is that flavouring? In terms of our concentric circles of context, the strength of the influence will depend on proximity: propitiatory ideas nearby (say in the same verse or chapter) will influence the sense of hilas- more, while those in the wider context (say, elsewhere in the same bible book) would be in an outer circle, and so flavour hilas- only slightly. In other words, even when the term means ‘cleansing’, the sense of propitiation can be overlaid on this to varying degrees.


This brief survey of hilas- in the LXX has yielded some helpful results. It has identified the ‘home ground’ of the group: Leviticus/Numbers, and its most common semantic territory – Levitical cleansing. It has also given us two outer limits of the word-group’s meanings: simple ‘cleansing’ and simple ‘propitiation’. We have also seen how these meanings can be combined and sit together in a single occurrence.

That’s our OT survey. Now we are ready for the main event: how is hilas- used in the NT?

Tomorrow: hilas- in the New Testament

confess(In the previous post we began our assessment of the work of Leon Morris on the word ‘propitiation’. Here’s the rest.)

Next Morris turns to hilasmos. He concludes once again that ‘whenever’ it means forgiveness the circumstances indicate the turning away of divine wrath.

One example should be enough to test the quality of this conclusion. In Psalm 130, we read ‘but there is forgiveness with you’. The context celebrates God’s loving orientation towards Israel. He does not mark their sins but will instead redeem them from them all. It would seem that God’s wrath is far from view in this Psalm. But no, Morris finds propitiation implied even here: “the word occurs in a context of trouble.” That’s it, that’s the proof. Did you catch it? The writer is facing troubles: a clear statement of God’s wrath(!) Apparently God has marked the psalmist’s sins after all. And not only is God’s wrath here: for Morris the turning away of wrath is a ‘necessary feature’ implied by this context of trouble.

It seems there is no context where Morris cannot find the idea of God’s wrath being turned away. I suppose he might say the whole Bible is such a context – in which case every word in it must necessarily speak of propitiation! In any case, Morris’s approach makes the business of considering individual examples of usage pretty pointless: we know what the outcome will always be.

Next is hileos. Morris’s comment on Deuteronomy 21:8 is revealing. The context is all about the removal of guilt in case of murder. Clearly expiation is in view. The elders are to sacrifice a heifer and say, “Do not let the guilt of innocent blood remain upon your people.” There is no mention here of wrath or punishment. In spite of this, Morris detects it: “It is difficult to interpret this other than as a propitiatory rite.” Why is that? Why could it not be, say, an expiatory rite? No reasons are given. In fact we already know that Morris has ruled out the category of simple expiation a priori. Actually one gets the feeling that this comment of Morris’s pretty much sums up his approach to studying the usage of the hilas– word-group: he finds it difficult to seriously consider interpretations other than propitiatory ones. If so, this tells us about Morris – but not about hileos.

The real test for Morris comes with exilaskomai.  For the word is mainly used in Leviticus, in a cultic context. But in Leviticus there is no mention, no suggestion of God’s wrath: the focus is always on the removal of sin or guilt. In fact the book is remarkable for omitting this common OT idea, the wrath of God. In this cultic sphere, the category is apparently absent.

You might think it would be difficult to find propitiatory ideas, then, in this Levitical term, exilaskomai. But I think by now you’ll have guessed that Morris finds them. Here’s how he does it: there are a handful of occurrences of exilaskomai outside this cultic context. He decides that this reflects the normal usage of the word, and that the huge number of cultic occurances are the exceptional usage. Ok…

He calls these few non-cultic occurrences “what the verb means in itself quite apart from the conventional use of the cultus.” He then demonstrates to his own satisfaction that these few occurrences allow of a propitiatory interpretation. The next step is to let these few non-cultic occurrences to control the meaning of the term when found in a cultic context: they ‘give us the key to the understanding of the cultic references.’ They become in effect the tail that wags the dog. They force a propitiatory meaning onto the Levitical usage.

This is Morris at his worst. As we have pointed out elsewhere, this idea of an intrinsic meaning in a word ‘in and of itself’ apart from context and usage, this has been thoroughly discredited by a century of lexical semantics by now. Usage and context are the keys to word meaning. For Morris to extract the word from its cultic setting in order to pin down its ‘real’ meaning apart from all the distractions of context, is naive. To then reimport this meaning into the cultic usage, is to commit the ‘illegitimate identity transfer’ which James Barr complained was such a common error among biblical scholars. A word’s meaning in one context cannot determine its meaning in a very different context.

In fact, words take on different meanings in different settings. Especially in a technical setting like the Levitical instructions, a word could easily have a special meaning. The legitimate way to discover that meaning is to look at Levitical usage. Simple. But that yields an ‘expiation’ result…

At the end of this chapter studying the OT usage of the hilas- word-group, Morris summarises his position beautifully: “When we reach the stage where we must say ‘When the LXX translators used “propitiation” they did not mean propitiation’, it is surely time to call a halt. No sensible man uses one word when he means another.” This would seem to be a conclusive argument – if the LXX was using ‘propitiation’. But that is precisely what Morris is attempting to prove. What this sums up so clearly is that Morris has all along been assuming the meaning of the word in order to prove that assumption. He might as well say, “It must mean propitiation, because that’s what it means.” We might remember that he began his discussion with this same argument. This silliness does nothing to advance our understanding of the hilas- group.

Such large-scale and persistent methodological flaws and follies as we have identified render Morris’s work of little value as a contribution to the study of the hilas word-group. The fact is that neither he, nor the ‘authorities’ he adduces, seem to understood how to employ the disciplines of modern lexical semantics (the science of studying words).

Sadly the evangelical constituency for which he was writing had even less understanding of these things than he did, and so were easily impressed by the appearance of scholarship. I for one grew up on this diet, being assured that “it has now been settled by the best scholars that this word means ‘propitiation’”. It is distressing to revisit this from a linguistic point of view, and find such poor quality work. I come away from this review feeling that the Christian community deserved better from a scholar they trusted so much.

In summary, pretty much everyone agrees that the idea of ‘expiation’ is central to in LXX usage of the hilas- word-group. Morris has argued that propitiatory ideas are also to the fore whenever this root is used. But he has not produced convincing evidence for this. The best we could say is, he has shown from the usage that ‘propitiation’ is sometimes in view in some of the terms studied. But in other hilas- words – in particular those connected with the Levitical sacrifices – there is no evidence of ‘propitiation’.