If something is scalable that means that the project is able to operate as it gets bigger.

What's the inverse of this? Suppose I have something that is big, but it can be broken down into subsections and it will still be able to be broken down, etc. — something like a fractal. That is, as it's broken down, it still contains the main components.

For example, I have a list that contains 16 items that can be broken down into 4 main categories, and each category can be broken down further into the same 4 categories. I'm looking for a word to describe this.

  • Firstly the question begins with a fundamental misconception: "scalable ...means that the project is able to operate as it gets bigger" --Wrong. Next, this question could have been off-topic for scalable being a technical term (its meaning is different in the contexts of modeling, software architecture, fractals, and so on). Above all, the OP apparently has not looked back ever since asking the question, showing no response to answers and comments in all the months. – Kris Nov 23 '12 at 10:46

11 Answers 11


First of all, you're asking for the inverse, not the opposite. (Inverse = reversed in direction, position, order, or tendency.)

Secondly, scalable works in both directions. If you want to be specific about it, you can say it "scales up" or that it "scales down". For an adjective, you can say that it is "scalable downward".

  • +1 Yes, that's the basic point here. Things can scale -- upward or downward, not necessarily upward only. – Kris Nov 23 '12 at 10:41

One adjective that may fit the situation you describe is scale-invariant (i.e. no matter how big something gets you still have to approach it in the same exact way as you would a smaller version). This is akin to self-similar, a typical descriptor for a fractal.


You could try limited. Many computer systems can be scaled to a degree until they hit some limit (such as memory, network bandwidth, processing capacity, etc.).

However, the process you describe (dividing a list into categories) is called decomposing, not scaling.

  • If I were in a meeting discussing whether or not some new project should be scalable, I'd certainly be likely to use the word limited to describe the alternative. The fact that OP isn't clear on what 'scalability' means is all the evidence we need to suggest that one should be wary of using opaque terminology. Limited is accurate enough for the context, and should be easily understood. – FumbleFingers Jul 14 '11 at 3:17

From a business perspective, the opposite of scalable is implosive. A scalable business will have its net profit grow along with the growth of its customer base. However an implosive business will only be profitable with a small customer base and will have its profit diminish when it adds on more customers. Thus the business implodes under the weight of its customer base.

A scalable business enjoys the economies of scale and thus its profit grows as its customer base increases. Preferably profit grows at a faster rate than the customer growth. A scalable business

However an implosive business' marginal profit decreases as customers are added, which reduces the overall profit as the business gains more customers. Therefore when the customer have exceeded a certain number, the business could no longer be profitable and implodes.

An implosive business


Modular would probably work in most contexts.


While I disagree that "scalable" has a single direction, even in computing, "granulate" or "granularize" might be the word you want. See also "granular computing".

  • So granularizable or granulatable? Googling, granularize is the verb used for this meaning in computation, whereas granulate means to turn a physical material into granules. So the right word is granularizable. – Peter Shor Nov 23 '12 at 15:00
  • I'd say Granular. – SF. Nov 23 '12 at 15:09

Some possibilities are reducible, tractable (in computer science), decomposable, divisible, and self-similar (fractal-like).

Also, to me, what you want isn't the opposite of scalable. I'd say something that can be reduced is still scalable.

  • Tractable problems are simply all those which are capable of being handled by computational methods, given realistic demands on processing power. – FumbleFingers Jul 14 '11 at 3:19
  • A scalable solution is normally by its very nature decomposable, because as the demands grow you may have to replace different 'original' components with better ones that either improve throughput or handle more variations. So you need to decompose the solution into those replaceable modules right from the start. – FumbleFingers Jul 14 '11 at 3:24
  • I chose "divisible" but found it here already. – GEdgar Nov 23 '12 at 17:09

Sounds like you're describing another property of fractals — self-similarity:

An object is said to be self-similar if it looks "roughly" the same on any scale.

So the list you're describing is self-similar in that you get a repetition of the same categories at different levels of 'scale' ("16 items can be broken down into 4 main categories, and each category can be broken down further into the same 4 categories").


The word I would use is shrinkable.

Scalable implies "the bigger the better". Shrinkable implies "the smaller the better", especially if you are trying to fit more things into, say, a suitcase.


I would use the word "elastic". In economics the word is used to denote demand that highly depends on price in both directions.

  • You are expected to also say why you would think so; what references you can site in support of the contention; any precedents you can show as in published literature. As a not-so-new member, you are aware of this. – Kris Nov 23 '12 at 10:50
  • @Kris in economics the word is used to denote demand that highly depends on price in both directions. – Anixx Nov 23 '12 at 14:32
  • That is a good point to include in the answer. I'll even withdraw my comment if you do. :) – Kris Nov 23 '12 at 14:34

May I suggest aggregative? Being aggregative reverses to quality of being divisible.

Not the answer you're looking for? Browse other questions tagged or ask your own question.