Functional Why-Why Analysis
Aleksey M. Pinyayev
In late 80’s , I started using why-why analysis for decomposing complex inventive
situations into the simple problem statements. Later I have discovered similar approach in
the research of Kishinev school – the work that led to the development of ARIZ-SMVA, the most detailed version of the famous algorithm. Today, why-why analysis is widely
used in TRIZ for better understanding of sophisticated inventive problems. The essence
of why-why analysis is simple – keep asking “why” until you find out the causes of the problem. Years of experience with this simple but powerful technique led me to its better
understanding and development. Why-why became much more universal and robust tool,
not only for problem definition but also for coming up with creative options. It also
integrates seamlessly with my recently developed system of Functional Clues .
The Myth of the “Right Problem”
The days when TRIZ practitioners accepted the problem statement suggested by the
problem owner are long gone. An understanding that the problem needs to be re-defined
in order to be successfully solved is now a commonplace. In this new and better world of
understanding the importance of problem analysis, a new myth has been created – a myth
of the “right problem”. “Work on a Right Problem” is a widespread motto in engineering
communities and especially in TRIZ. The assumption here is that the Right Problem can
somehow be found before the problem is solved and that its solution will satisfy the
initial inventive situation in the best possible way. However, it is very easy to point out to
the right problem after the fact, when it has been solved and delivered this best possible
solution. Try defining the “right problem” before getting solutions, and it will be a very
difficult or impossible task, simply because there are no criteria that dissimilate “right”
and “wrong” problems.
My approach is to break the initial inventive situation down to a multitude of
problems and solving most if not all of them. The outcome of this work is a number of
solution options which are prioritized based on a set of acceptance criteria. These criteria
reflect the reality of the project: they define the most practical way of reducing the
concepts into practice. Indeed, there are much more problems to solve with this approach,
but, in fact, this makes the invention work easier because it expands the vision of the
inventor and allows to look at the different sides of the problem at once.
The case study I will use is decaffeination process. This is a real technical problem
which has been analyzed and solved at P&G. The main concept coming out of this
problem solving process became a cornerstone of one of the major P&G plants –
Sherman-Texas Decaffeination Facility. I will begin with a brief description of the
traditional decaffeination process which was the most known and widely used process
before the new technology was invented.
1Traditional Decaffeination Process
The simplified schematic of the traditional decaffeination process is shown in Fig. 1.
This batch process uses organic solvent – ethyl acetate with some water – in order to extract caffeine from a bed of coffee beans. In the process, the solvent is supplied at the
top of a large tank with coffee beans and collected at the bottom of this tank. The second
step of this process, called distillation, separates caffeine from the solvent and returns the
solvent back into the main tank. Decaffeination process is followed by the steam
extraction which removes the residual solvent from the beans. Leaching takes 20 – 24
hours per batch followed by 24 hours of the steam extraction. Because the process is so
slow, the batch size has to be very large, which means high capital cost. The objective of
the problem analysis and solution process was capital reduction.
96% ethyl acetate
4% water solvent
beans LEACHING DISTILLATION
solvent + caffeine caffeine
Fig. 1. Traditional Decaffeination Process Paradox Is the Problem
The why-why diagram of the problem described above is shown in Fig. 2. The
diamond-shaped box is used for the initial observation and the magenta color is used to
show the ends of the cause-and-effect chains. The rules on where to end these chains are
described in the next section. As one can see, the why-why investigates three main
branches of causes related to the solvent, the bean and the interaction between them. The
why-why is build in layers, and a next layer is only built after a previous one is
completed. This means that all whys for a particular cause must be exhausted before the
causes of these new whys are identified. Finding hidden whys is an important objective
of why-why analysis. The technique for finding hidden whys is described below in the
section called “Hidden Why-Why Treasures”. The rest of the current section will explain
1 Courtesy of Lowen R. Morrison, P&G
Fig. 2. The why-why diagram of the traditional decaffeination process.
how the specific problem statements called “why-why contradictions” are derived from the results of why-why analysis.
The technique of why-why contradictions is based on the undesirable action negation
approach suggested by A. I. Ponomarenko  and further develops this approach into a
comprehensive problem analysis strategy.
Why-why contradictions (specific problem statements) can be derived from the
results of analysis by using any two consecutive whys. To form a contradiction, we keep
former why and turn latter why around. For example, let’s consider a chain which ends with why # 37. The contradiction between whys ## 37 and 36 will be this: how to add
more water into the bean without changing the bean’s flavor? The water content in the bean is an optimum between extraction and flavor impact. Can the beans be pre-treated
for better flavor retention? What are the ways of flavor recovery after water addition?
How water pH influences flavor impact? These are some of the possible ways to resolve
this contradiction. The contradiction between 36 and 12 will be as follows: how to fully
solubilize caffeine inside the bean without adding more water into the bean? This is a
problem of changing solubility of caffeine in water. The next contradiction in this chain
is between 4 and 12: how to optimize caffeine availability in the bean without fully
solubilizing the caffeine in it? Can we, for example, rupture the walls of some of the cells
without breaking the bean? This would increase caffeine availability without additional
A. I. Ponomarenko suggests applying undesirable action negation to all whys in a
chain and then solving all problems obtained by such negation. I add to it using a concept
of why-why contradiction which combines negated why with the previous why, building
a comprehensive list of all why-why contradictions (Excel is a good tool to use for that)
and prioritizing these contradictions based on a set of acceptance criteria before these
contradictions are resolved. The acceptance criteria may include the magnitude of the
benefit coming from a potential solution, novelty, level of system modifications required
by a potential solution, availability of the substance-and-field resources, safety,
knowledge availability and others. The key point is: why-why contradictions are specific enough in order to provide a lot of information about potential solutions
even before the problem is actually solved. For the set of three problems considered above, some are well-known and difficult (changing solubility of caffeine in water),
others look novel and promising (rupturing some of the cell walls without breaking the
bean). It is apparent, for example, that the problem of avoiding flavor impact requires
significant research with no clear perspective on getting a good solution. All these
considerations are used in order to prioritize the contradictions. I use the name Why-Why
Contradiction Map for the result of why-why analysis process – prioritized list of why-why contradictions. The Contradiction Map allows to approach problem analysis
systematically and avoid missing the key contradiction. Obviously, the problems are
solved in the order of their priority which in most cases allows to significantly reduce the
amount of the required work.
When to Stop Asking “Why”
It is important to define the boundaries of why-why analysis in order to limit
redundant work. It is also important to know when your why-why analysis is complete.
Additionally, it is useful to fully understand the limitations of the project. This section
provides specific respective techniques.
I recommend doing a first cut definition of the project limitations as a pre-requisite
for why-why analysis. These limitations will be used to define the analysis boundaries.
Let’s refer again to the decaffeination process. This is the first cut of the project
limitations defined at the project definition stage:
? Ethyl acetate must be used as a solvent
? Must use the same raw beans
? Cannot increase capital cost
Every time we add a why, we will do a turnaround of this why and compare it with
the project limitations. If the turnaround why goes against the limitations, we will cut our
why-why. Let’s consider the branch 3 – 10 – 21. The diffusion of the caffeine out of the
bean is too slow because the solvent is not optimal. This is the turnaround why: “the
solvent is optimal”. If we go this way, our problem is how to optimize the solvent. This problem does not necessarily go against any of the limitations listed above. Parameters of
the ethyl acetate such as concentration, viscosity and temperature can all be optimized for
the process. However, one of the reasons why ethyl acetate is not optimal is its limited
extraction capacity. This capacity, basically, defines the solvent and is an intrinsic
property which differentiates given solvent from other solvents in its class at the same
temperature and concentration. The turnaround why is this: “extraction capacity of the solvent is high”. This turnaround why, which leads to a different kind of solvent, violates
the requirement to use ethyl acetate. At this point, we cut our why-why for this branch. I
use color-coding to mark the root whys as shown in Fig. 2.
It is important to see that while we cut off the branch, we still include the root
contradiction into the Problem Map. The root contradiction for the branch 3 – 10 – 21 is
“how to optimize the extraction effectiveness of the ethyl acetate without changing its
extraction capacity”. This is a valid problem which does not violate the project
constraints and it will be included into the Contradiction Map.
The why-why analysis is complete when all of its branches are cut off. In some
special cases, turnaround whys may become solutions and this also cuts off the branch.
Let’s consider the branch ending with why # 24. The turnaround why for it is “high
pressure does not increase the capital cost”. Normally, the capital cost of the high-
pressure parts is defined by the industry average which cannot be changed, so it becomes
a project limitation. However, if there is a low-cost supplier of the quality pressure parts,
this may become one of the solutions. Either way, the branch is cut off.
This technique also helps to make the project constraints list more compete. Let’s
consider root why # 37. The turnaround for it is “too much water does not impact the
flavor” which does not violate any of the project limitations in the original list. However,
the problem of avoiding the off-flavor caused by excessive water may go beyond the
scope of our particular project. In this case, we add “too much water impacts the flavor” into the list of constraints for the project. It is important to understand that the why-why
boundaries are project-dependent and what can be done in one project may be
unacceptable for another.
Hidden Why-Why Treasures
In many cases, one can find hidden whys between apparent ones which are already in
the diagram. A. I. Ponomarenko  mentions an excellent example which can illustrate
this approach. A. I. Ponomarenko provides this example without any analytical technique
or tools to help in such analysis. The example describes the crashes and breakdowns of
the high-speed power machines such as compressors, steam and gas turbines which use
blades to convert kinetic energy of the gaseous media into rotation of the main rotor.
Blade break-off leads to rotor brushing the stator with subsequent machine crash. A. I.
Ponomarenko’s observation goes as follows: “The blade break-off has already happened;
however, the sudden misbalance of the rotor has not happened yet. This time lapse is very
small, but it exists, and one can formulate a problem with a corresponding Ideal Final
Let’s step back from this involving example and think about it from the methodical
standpoint. Initial, apparent chain of events went like this: insufficient blade strength ?
blade break-off ? rotor touches the stator ? machine crash. However, there is a hidden
why between “blade break-off” and “rotor touches the stator”. This hidden why is “rotor
imbalance”. Rotor imbalance happens after the blade brake-off but before the rotor began
to touch the stator. Understanding this hidden why leads to a very unobvious and new
why-why contradiction: how to eliminate rotor imbalance after blade break-off but
before the rotor begins to touch the stator? Because this problem is not obvious, chances of getting novel and powerful solutions are very high. A. I. Ponomarenko mentions ring
auto-balancer as a potential solution and there may be more options.
I have found that discovering the hidden whys is a very powerful tool in re-defining
the problem. Initially, my technique was very simple but not very instrumental: look at
the two consecutive whys and try to complete the sentence “The first leads to … which
leads to the second”. Staring for a while into the empty space between the first and second whys helped to come up with the hidden one. Later, I noticed that most of the
hidden whys were related to either analysis of the sequence of events in time or
considering the process on a micro-level. The example above demonstrates the value of
the sequential analysis: what happens when the blade has already broken off but the rotor
has not started to touch the stator? With this “time zooming”, finding hidden why is
easier. The example below demonstrates the micro-level approach.
Procter and Gamble makes the product called Bounce for more than 25 years. Bounce
is a nonwoven sheet which is impregnated with the actives that make clothes softer and
smelling fresher after drying them in a dryer. Bounce sheet tumbles in a dryer along with
the clothes; the actives get transferred to the clothes by the mechanical motion and
friction combined with elevated temperature. A dryer so common in American homes
today is a machine which passes hot air through a load of clothes which are continuously
tumbled by the machine’s internal rotating drum. The problem I did why-why analysis for was how to avoid wasteful perfume removal by the hot air in a dryer. The perfume,
which is one of the Bounce’s actives, can only be transferred via direct contact with the
clothes. When it volatilizes, it is gone with the air blowing through the dryer. My initial
why-why chain looked like this: air inside the dryer is hot ? perfume is exposed to hot air ? perfume heats up ? perfume evaporates ? air removes perfume molecules. In
order to find hidden whys, I asked myself a question: when these transformations occur,
what happens on a micro-level? Let’s use this approach to find a hidden why between
“perfume heats up” and “perfume evaporates”. How does the evaporation happen on a
micro-level? Perfume in the Bounce sheet is incorporated into the cyclodextrine
molecules. This enables a long-lasting fragrance effect: cyclodextrine molecules are
small “cages”, each holding one perfume molecule inside. These micro-cages attach and
hold perfume on the clothes for a long time after removal from the dryer, gradually
releasing precious fragrance into the air. In that context, evaporation means that excited
perfume molecules vibrate so strong that this vibration pushes them out of their
cyclodextrine cages. The new, hidden contradiction coming from this analysis was this:
how to prevent jumping of the perfume molecule from the cyclodextrine shell in spite of
the molecules being hot and vibrating? This new contradiction led to a few novel
To recap, the technique for finding hidden whys looks like this:
? Define two consecutive whys
? Build the time diagram of the events described by the two consecutive whys
? Answer the question: what happens when the first event has completed and the
second one has not yet started? Right down hidden why.
? Define first and second events on a micro-level
? Answer the question: how the transition from the first to the second event happens
on that level? Right down hidden why.
Obviously, doing this for all whys is a lot of work, but this work can be reduced. I found that finding hidden whys is most critical for the whys closest to the roots of the
Now we will come back to the decaffeination process and see if we can find a hidden why between the whys ## 20 and 11. The bed of beans has high resistance to the solvent
flow and this leads to something which leads to the solvent not getting replenished
quickly. We can easily see that the time domain analysis is not going to help very much
in finding the hidden why because both consecutive whys happen at the same time. Let us
try the micro-level approach.
The first micro-level below the bed of beans is a few individual beans with the solvent surrounding them. These beans are as close as to touch one another, so it is
difficult for the solvent to go around the bean quickly and this is why the solvent is not
getting replenished quickly! Our hidden why is low flow rate around the bean and our
new why-why contradiction is how to increase the flow rate around the bean without
changing the bed resistance to the flow. Now, it is simply an Ohm’s law: in order to
increase the current, we must increase the differential pressure. Now we no longer rely on
the gravity to create this pressure differential but more on the well-known methods of
maintaining high pressure differential through the resistive media such as pipes and
pumps. We replace the huge tank with a sequence of bean-filled pipes and we maintain
high flow rate around this system of pipes by applying appropriate pressure differential.
This is the breakthrough concept  around which the Sherman-Texas plant was built.
Functional Why-Why Analysis
Usually, we do functional analysis for the entire initial situation to help us better understand the problem and define the problem model (step 1.2 and 1.3 of ARIZ-85V
). My approach is different. In this approach, we do the functional analysis of the why-
why contradictions as the next step to the completion of the why-why diagram. We end up doing many small functional analyses instead of a big one. On the level of why-why contradiction, the functional model is only slightly more complex than the problem model. This approach allows to make the functional analysis much simpler and more focused which makes the concept generation easier and more effective. Additionally, this approach allows us to naturally transition from why-why analysis to Functional Clues . Now, let us discuss how Functional Why-Why analysis is done in TechOptimizer.
TechOptimizer v. 3.5 offers an interface very suitable for the Functional Why-Why
analysis. This analysis can be done in the Process Analysis module (Fig. A1.1). Please, refer to the series of the screenshots in Appendix 1. Importantly, Show Subprocess stage feature must be enabled (Fig. A1.2). The why-why analysis is done in the Subprocess window of the software. The initial observation uses Not Analyzed Subprocess box and the whys are put in the Subprocess boxes. The whys are connected with the causal arrows. Please, refer to Fig. A1.3. Each Subprocess box has an option of opening the Operations window, separate for each Subprocess. This window is used to right down the why-why contradiction. This contradiction is described in the Operation box (Fig. A1.4). Each why has its unique contradiction. The Operation box has an option to open the Operation Analysis window. This window is used for the functional analysis of the why-why contradiction (Fig. A1.5). The whole why-why diagram will be saved along with all the contradictions and partial functional analyses. Interface with Functional Clues
There is only one step left which connects the Functional Why-Why analysis with the
system of the Functional Clues described in : the transition from the partial functional model to the Application Condition of a Functional Clue. This transition is done in a functional diagram by selecting one of its components and no more than two arrows coming to or from this component. This selection, basically, defines the graphical model of the problem also called Application Condition in the Functional Clue language. The  considers this transition and the following steps in details. The system of Functional Clues completes the problem analysis and solution method described here by providing means to generate creative concepts addressing each of the why-why contradictions. Conclusions
The Functional Why-Why approach described in the present paper allows to break the initial inventive situation down to the partial problem statements called why-why contradictions. These why-why contradictions are prioritized into a Why-Why Contradiction Map and further analyzed using functional approach. The initial why-why diagram gets further developed by the hidden why search. This systematic approach enables reliable problem analysis and ensures that the key contradictions or problems will not be overlooked. This, in turn, increases reliability, productivity and effectiveness of the concept generation stage done by the system of Functional Clues.
1. A. M. Pinyayev (1998). Causal Analysis of Undesirable Effects in Complex
Technical Systems. Collected articles: “Teoriya I praktika obucheniya
tekhnicheskomu tvorchestvu”, 21 – 27 May 1988, Chelyabinsk, pp. 65 – 66. In
2. A. I. Ponomarenko (1995). Selecting a Problem by Using the Undesirable Action
Negation Statement. Journal of TRIZ, 1995, ?1, pp. 51 – 53. In Russian. 3. B. L. Zlotin, A. V. Zusman (1991). Come to the Training Ground. In collect.: How to
Become a Heretic. Compil. By A.B. Seljutskiy. Petrozavodsk, Kareliya, 1991, p. 192.
4. U.S. Pat. # 4,474,821
5. Aleksey Pinyayev. Functional Clues. TRIZ Journal (www.triz-journal.com),
Appendix 1. Using TechOptimizer v. 3.5 for the
Functional Why-Why Analysis.
Fig. A1.1. Functional Why-Why Analysis is done in the Process Analysis module of the