by Robert R.

Sterling

Ibward a Science
of Accounting
We accountants do not resolve issues, we abandon debated in the literature.' Such issues are periodi-
them. I do not mean to imply that we ignore issues. cally brought up for debate, abandoned for a while
Quite the contrary, we debate them long and loud. and then resurrected and debated again. Account-
However, the debate, instead of coming to a reso- ants anticipated the ecology movement by some
lution, continues until another issue comes along years; instead of disposing of issues, we recycle
that is more current and more controversial, and them.
then we forget the former issue. Consider the Instead of resolving issues in accounting, we
LIFO-FIFO debate that occupied so much of our move from one unresolved issue to another unre-
attention and journal space a few years ago. Cur- solved issue. Our inventory of unresolved issues
rently LIFO versus FIFO is a dead issue. Note that grows larger and larger. Additions to the stock of
I said "dead issue"; I did not say "resolved issue." unresolved issues arise in the natural course of
Despite the fact that it is not resolved, at the pres- events, but it has been virtually impossible for us
ent time there is no attention or journal space de- to subtract from the stock. Why can't we resolve
voted to it. The reason for this is that other more these issues?
pressing issues came along to divert our attention. The explanation for our inability to resolve is-
One can run through the list of unresolved issues sues is to be found in the way we conceive of the
and discover much the same thing: tax allocation, issues. We conceive of the issues in such a way that
accelerated versus straight-line depreciation, they are in principle unresolvable. The reason we
pooling versus purchasing treatment of business cannot answer our questions is because we phrase
combinations, full versus direct costing, etc. For the questions in a way that prohibits obtaining an-
example, the reason we quit debating full versus di- swers. We define our problems so that the very def-
rect costing was not because the issue was resolved inition precludes the possibility of a solution.
but rather because the merger movement caused us
Art versus Science
to divert our attention to the more pressing issue of
purchases versus pooling. Our misconceptions start with the broadest of is-
Further evidence of our inability to resolve is- sues—the definition of accounting. In the official
sues is the fact that we periodically resurrect dead pronouncements of our societies,^ in journal ar-
issues. The classification of bad debts and the ticles,' as well as in textbooks," we find accounting
treatment of interest as a cost were controversial defined as an "art" as opposed to a "science."
issues when I was a student. Conclusive evidence These authors contrast the methods of accounting
that these issues were not resolved at that time is with those of science and claim that accounting is
provided by the fact that they are once again being necessarily based upon conventions, as opposed to
the laws that form the basis of a science.
The definition of accounting is based upon two
Robert R. Sterling is Jesse H. Jones Frofessor, De- misconceptions. First, the authors have a miscon-
partment of Economics and Accounting, Rice Uni-
ception of the nature of science. They seem to be-
versity. This article is based on a presentation at the
Frice Waterhouse Distinguished Lecture Series, Stan -
ford University, April II, 1975. 1. Footnotes appear at end of article.

28 • FINANCIAL ANALYSTS JOURNAL / SEPTEMBER-OCTOBER 1975

lieve that in order for a discipline to be scientific it of uncertainties, e.g., the cause of cancer, but he
must consist of immutable laws and absolute does not fall into the trap of defining the problems
truths. The fact is that in even the most exact so that they are in principle unresolvable.
sciences laws are mutable. Witness the recent over- I think the analogy fits accounting. There is
turn of the Newtonian laws of physics by Einstein. nothing about our subject matter that requires that
To a scientist, laws are generalizations that must be it be based upon conventions instead of laws. It is
continually subject to empirical challenge. If they based upon conventions for the simple reason that
were immutable, scientists would call them "defini- we define it that way. It is possible to define it in
tions," not "laws."' some other way. There is nothing inherently un-
In direct contrast to scientists, accountants con- scientific about accounting; it is our approach to
ceive of laws as immutable and seem to believe accounting that has been unscientific. It is possible
there are no uncertainties in science. For example, for us to adopt a scientific approach.
Lawler writes:^
Objectives: Artistic versus Scientific
If, in truth, a practitioner of the accounting art is
a truthseeker, he is confronted by a host of un- Although I have never found the notion of ac-
certainties. There are no natural laws of ac- counting associated with science in accounting
counting which he can employ, only conventions texts, I often find such associations in science and
validated by experience, estimates born of philosophy texts. In fact, many definitions of
human judgments, guidelines developed by his science include "account for" or "account of" as
professional societies. an integral part of the definition of science. For ex-
ample, Torgerson writes:'
We can all agree that there is a "host of uncertain-
ties" in accounting. I believe it is the existence of The principal objective of a science, other than
such uncertainties that causes accountants to de- the description of empirical phenomena, is to es-
fine accounting as an art. If one conceives of tablish, through laws and theories, general prin-
science as consisting of absolute truths, then one ciples by hieans of which the empirical phenom-
must conclude that accounting is not a science ena can be explained, accounted for, and
since there is so much uncertainty in accounting. predicted, (emphasis added)
However, that conclusion is based upon a miscon- Insofar as I am concerned, this is also the objective
ception of the nature of science. There is also un- of accounting. Certainly we want to account for
certainty in science, else there would be no reason empirical phenomena. Hopefully, our financial
for scientists to continue their research. Thus, statements provide explanations to the reader. I
when accountants define accounting as an art be- also believe that we should seek laws which will
cause of its uncertainty, they are basing their defi- allow us to predict in the scientific sense. (Whether
nition upon a misconception of science. or not we want to predict in the sense of fore-
The second misconception is that there is some- casting is one of those current unresolved issues I
thing about accounting that makes it inherently want to avoid in this article.)
unscientific. That accounting is based upon con- Let us contrast the objectives of art to those of
ventions instead of laws is the present condition of science. The purpose of an artist is to present his
accounting, but it is not a necessary condition. personal "interpretation" of an object, not to pre-
Compare the primitive medicine man to the sent an accurate representation of that object. Thus
modern medical scientist. Both study illness—they the picture that shows up on the canvas is due
have the same subject matter—but they differ in more to the imagination of the artist than to the
their approach to that subject matter. One per- object he is painting. Artists (at least post-Renais-
forms artistic rituals designed to exorcise demons. sance artists) strive to be different—to present dif-
The other performs the arduous scientific task of ferent depictions of the same object. For example,
looking for empirical generalizations or laws. Thus if Picasso had painted a picture of Madame
the difference between them is not the nature of Matisse, he may have depicted her with only one
the subject but rather their approach to that sub- eye since there are in fact Picasso paintings
ject. Once the medicine man defines illness as the showing one-eyed women. On the other hand, as
product of the inexplicable caprice of demons, he you will recall, Matisse depicted her with two
is trapped. His problems will remain unresolved eyes—and a green nose. It is unlikely that Picasso
because he has defined them in such a way as to would have chosen green as the color of her nose.
make them in principle unresolvable. The first step Both Picasso and Matisse would be looking at the
toward science is to change the definition of the same object, but the representations that would
problem. The modern medical scientist faces a host show up on the canvas would be quite different. In

FINANCIAL ANALYSTS JOURNAL/ SEPTEMBER-OCTOBER 1975 D 29

art, different representatiotis of the same object are human trait to want to be flattered, to be shown in
not only acceptable, they are highly desirable. The the most favorable light. Since managements are
representations are intended to depend upon the human, I cannot imagine that they instruct their
imaginations of the artists, not upon the object. accountants to "paint me as I am, warts and all." I
By contrast, scientists strive to avoid such differ- suspect that they give opposite instructions:
ences. If two scientists were to count the number of Flatter me just a little—a 10 per cent increase in
Madame Matisse's eyes or determine the color of earnings per share would be about right. Conceal
her nose, differences between them would be cause my warts and pimples with cosmetically phrased
for alarm. The same is true if they were to count footnotes.
her white blood cells or weigh her or diagnose her
ailment. A basic tenet of science is that there be Since most of us are not as blunt as Cromwell, I
agreement among independent observers. The rep- omitted the threat to never pay a farthing if the in-
resentations are intended to depend upon the ob- structions were not followed. Nonetheless, I expect
ject, not upon the imagination of the scientists. that the threat is there, albeit tacit. The controller
or auditor is keen enough to recognize that he is
Defining accounting as an art is, unfortunately, a
likely to be replaced if he does not follow in-
fairly accurate description of the current condition.
structions.
There is that infamous memo in the Penn Central
Case that says that one person's "imaginative ac- The danger is that, if we define accounting as an
counting is adding millions of dollars annually to art, then we leave ourselves open to receiving such
our reported income."' Certainly in that case the instructions and threats. If we defined accounting
representation depended more upon the imagina- as a science, we would be considerably less
tion of the accountant than upon the fmancial po- vulnerable.
sition of Penn Central. I fear that there are a good
Conventions versus Laws
many such instances. In the present unscientific
state of accounting, I fear that the representation If accounting is to be an empirical science, then
may depend more upon the imaginations of the ac- we must redefine our discipline so that it is based
countants than upon the firms being accounted for. upon laws instead of conventions.'" Of course, all
In this sense accounting may be in fact more like sciences have need for some conventions, but these
an art than a science. are relatively unimportant matters. For example,
This is not the way that things ought to be, how- an inch is defmed by convention to be a certain
ever. If two accountants were to count Madame length. It could have just as easily been defined as
Matisse's eyes, a difference in the count should be some other length, and that fact is what makes it a
cause for alarm. The same is true if they were to relatively unimportant matter. The selection of a
measure her income or assets or liabilities. A basic particular convention is an arbitrary choice.
tenet of accounting ought to be agreement among Therefore, conventions are things that are decreed,
independent observers. In accounting the measure not debated.
ought to depend upon the condition of the firm, We have need for conventions in accounting, but
riot upon the imagination of the accountant. In we must be clear on the distinction between con-
short, the objectives of accounting ought to be ventions and laws. At the present time the texts de-
identical to the objectives of any of the sciences. fine accounting as being based upon conventions in
general and then talk about various "cost conven-
Instructions to Artists tions" in specific cases. For example, in regard to
inventory valuation, LIFO is described as one con-
Perhaps the danger of defining accounting as an
vention and FIFO as another. In regard to depreci-
art can be more vividly illustrated by considering ation, declining balance is described as one con-
Oliver Cromwell's instructions to his portrait pain- vention and straight line as another. Despite the
ter.s fact that these are described as conventions, the
Mr. Lely,..., I desire you would use all your texts present arguments about the relative merits of
skill to paint my picture truly like me and not each and then say that the choice in a particular
flatter me at all; but remark all these case "depends upon the circumstances."
roughnesses, pimples, warts and everything, Although arbitrary, conventions, once decreed,
otherwise I never will pay a farthing for it. must be followed consistently. Thus it is contradic-
The very fact that Cromwell could give these iri- tory to say that the selection of a particular con-
structions to his painter is significant. It means that vention depends upon circumstances." Consider
he could have given instructions to do the op- the convention of driving on the right side of the
posite. I suspect that it is an ajmost universal road in this country. That this is truly a convention

3 0 • FINANCIAL ANALYSTS JOURNAL/ SEPTEMBER-OCTOBER 1975

 is evidenced by the fact that they drive on the left income or depreciation impossible until a future
in England. Since it is a convention, there is no event occurs. This wreaks havoc with the system.
need to continue the debate about which is The definition prohibits resolution of disagree-
best—either is equally as good as the other. We ments. If one conceptually makes a present magni-
simply need to make a decision and then forget tude depend upon subsequent events, then the
about it. A similar convention exists in accounting. present magnitude can never be known because
We put the credits on the right. Since putting the subsequent events always lie in the future and the
credits on the left would be equally as good, it is future can never be known—it can only be
easily recognized to be a convention. There is no estimated or forecast. This notion is particularly
need to debate the issue. pernicious because it commits one to a continual
Conventions are decided by decree, by fiat. To correction of past magnitudes or else an admission
drive on the right side of the road was decided by a that all past magnitudes are wrong.
legislative body. If these various cost allocations Suppose we were to calculate my weight in the
are truly conventional, then we need to do the same way that we depreciate an asset. We know
same thing in accounting. Let Marshall go up to that at acquisition (birth) Sterling weighed 10
Mount Stamford and fast for 40 days and 40 nights pounds. I estimate that his salvage weight at time
and at the end of that time return with stone of disposal (death) will be 210 pounds and that his
tablets that say: useful life will be 50 years. On the basis of a
1. THOU SHALT USE LIFO; straight-line allocation, I would report his present
2. THOU SHALT USE STRAIGHT LINE; weight, at age 42, as 178 pounds. If another ac-
3. THOU SHALT FLOW THROUGH; countant were to estimate that Sterling's useful life
and so forth. If Marshall has any difficulty in will be 100 years, then he would report his present
deciding on which conventions to choose, I would weight as 94 pounds. There is no way to resolve
be glad to loan him my solid-state binary decision- this disagreement. The disagreement is in principle
maker—he can fiip my coin. unresolvable because we have defined it so that his
In choosing conventions the important thing is present weight depends upon h\% future life. Since
to select one and then be consistent—which con- we cannot resolve disagreements regarding his
vention is selected is not important. Consider again future, we cannot resolve disagreements about his
the convention of driving on the right side of the present weight.
road. If the legislative body let each individual This definition causes additional problems. Sup-
choose which side of the road to drive on, the pose that we have been reporting Sterling's weight
result would be chaos. If we let each individual (on the past balance sheets) and his weight change
choose which side to put the credits on, the result (on the past income statements) on the basis of a
would be chaos. We have let each individual 50-year life estimate. Suppose that present events
choose between LIFO and FIFO, accelerated and now cause us to believe that Sterling's useful life
straight line, etc., and the result has been chaos. estimate should be changed from 50 to 100 years.
The point is that selection of any particular con- What do we do? Do we correct all the past balance
vention is arbitrary. It does not make any differ- sheets and income statements or do we admit that
ence which convention is selected, but it is impor- all the past statements were wrong? If we decide to
tant to be consistent. In accounting we have correct all of the past statements, what good does it
defined the problem as the choice among alterna- do? That is, how much does it benefit a reader to
tive conventions, but then we have confounded discover in the 197O's that the financial statements
ourselves by making the selection of conventions of the 194O's are wrong? Although we can change
depend upon unspecified circumstances, thereby our past statements, the reader cannot change the
precluding resolution of the problem. past decisions made on the basis of those state-
ments. Also, if we adopt the policy of correcting
Measurements versus Allocations past statements, then we will need to correct all the
In addition to defining the problem of deter- past statements again next year when next year's
mining costs and income as a choice among con- events cause us to change our estimates of his
ventions, we also define it so that a present magni- future life or salvage value again.
tude depends upon a future magnitude. For ex- Given these considerations, we will likely decide
ample, we say that the "true depreciation" cannot that correcting past statements is not a desirable
be known until the asset has been sold. We say that policy. But what is the alternative? Changing the
the "true income" cannot be determined until the estimate of my useful life from 50 to 100 years
firm has been liquidated. The problem here is that means that there is a change in my recorded weight
we have by definition made the determination of of 84 pounds, a write down from 178 to 94

FINANCIAL ANALYSTS JOURNAL / SEPTEMBER-OCTOBER 1975 O 3 1

pounds. Do we report that 84-pound weight loss until it came to rest. That is, if I dropped an object,
on the 1975 income statement? It seems odd, to then you would have to wait until it hit the floor
say the least, to report an 84 pound loss in 1975 before you could measure the distance traveled.
when one thing that we know for certain is that my My accounting ancestors just shrugged that off and
weight did not decrease by 84 pounds in that year. said it was too bad, but there was nothing that
Given these considerations, we will likely reject could be done about it.
this alternative. But we have already rejected the Luckily, I had other ancestors who decided to
other alternative. We have only two alternatives take that as a challenge instead of resigning them-
and we have rejected both: a classic example of the selves to the problem. They tried to develop
horns of a dilemma. methods of making observations of the object while
The dilemma is due to the way we define the it was still falling. In the early days these methods
problem. One is not required to forecast the future were crude observations of balls rolling down in-
in order to make a present measurement. In fact, clined planes. By continuing to work on the prob-
that is a contradiction of terms. Measurement is lem, scientists developed means of making precise
the process of discovering a present existing magni- measures. They now measure the distance traveled
tude. It is not a process of allocating past magni- by falling objects in evacuated tubes with high
tudes to time periods on the basis of forecast future speed photography.
magnitudes. In science, the determination of my We must admit that our present measurement
present weight would be defined as a measurement, problems are formidable. In the same way that the
not as an allocation. Defining it as a measurement ancients could not measure the distance traveled
does not automatically solve the problem, but it until the body came to rest, we cannot (so it is
does make it possible to solve it. Defining the alleged) measure the cost of a depreciable asset un-
problem as an allocation makes it in principle til it is sold. The difference is that we seem to have
unresolvable.''^ resigned ourselves to our inability to measure,
When I was a student this definition resulted in while the scientists viewed their inability as a chal-
an issue known as "all inclusive income versus lenge. The point is that we could also view our
dirty surplus." That the problem was never measurement problems as a challenge instead of
resolved is evidenced by the fact that we are still resigning ourselves to unresolvable problems.
debating it under the name of "extraordinary An Example of an Accounting Law
items." Today's controversial "big bath ac-
counting" issue is the same problem with a differ- In the previous discussion I said that we ac-
ent name. When a firm's management takes a "big countants have defined various cost allocation
bath," it changes past overly optimistic estimates of methods as being conventional. It is possible to
the firm's future to a present more pessimistic esti- define them as scientific laws—as testable empiri-
mate. This change of past optimistic estimates of cal generalizations. Consider depreciation. In their
the future to present pessimistic estimates of the discussion of the definition of depreciation, the
future results in changing past reported profits to Committee on Terminology says:'^
losses, permitting management to change future The term [depreciation] is broadly descriptive
losses to reported profits. As long as we define past of a type of process, not of an individual
and present measures to be based upon future mag- process, and only the characteristics which are
nitudes, such problems will remain unresolved. common to all processes of the type can
properly be reflected in a definition thereof.
The Challenge of Measurement These common characteristics are that a cost or
Somebody is sure to tell me that the situation is other basic value is allocated to accounting
unfortunate, but there is nothing that can be done periods by a rational and systematic method and
about it. The problem, they say, is that it is impos- that this method does not attempt to determine
sible to make the measurements that I suggest and, the sum allocated to an accounting period solely
therefore, we must resign ourselves to conventional by relation to occurrences within the period
allocations of past expenditures on the basis of es- which affect either the length of life or the
timates of the future. People have, in fact, told me monetary value of the property. Definitions are
that. That may be an appropriate artistic or ac- unacceptable which imply that depreciation for
counting response, but it would not be considered the year is a measurement, expressed in mone-
an appropriate scientific response. tary terms, of the physical deterioration within
Consider the following bit of (apocryphal) the year, or of the decline in monetary value
history: At one time in the past it was impossible to within the year, or, indeed, of anything that ac-
measure the distance traveled by a free falling body tually occurs within the year.

32 D FINANCIAL ANALYSTS JOURNAL/ SEPTEMBER-OCTOBER 1975

Let me repeat: "Definitions are unacceptable cerned with costs. I have been told this before.''' I
which imply that depreciation... is a measure- do not want to debate the merits of that issue here,
ment . . . of the physical deterioration... or the but rather merely to point out that accounting's
decline of monetary v a l u e . . . , or, indeed, of any- concern with costs is a consequence of the defini-
thing that actually occurs within the year." This tion of accounting. We could define it some other
definition precludes empirical tests because depre- way.
ciation is defined as a convention instead of some A more potentially damaging criticism of the
empirical measure of a testable phenomenon. proposed law is that historical cost depreciation
Let me redefine the problem. I want to propose may be a concept that has theoretical import even
an empirically testable generalization. That is, I though it lacks empirical import. In science there
want to propose a scientific law. First, in order for are two basic kinds of concepts: empirical and
it to be empirically testable, I must define an ob- theoretical. The empirical concepts must be sub-
servable. jected to empirical test. Theoretical concepts, on
Sterling's Definition: Depreciation is the decline the other hand, are subjected to logical test—the
in the exit value of produc- demonstration that the concept has some logical
tive assets. connections, via laws, with other concepts. In ac-
counting terms this means that a theoretical con-
Second, the law (actually a hypothesis until it is cept must be relevant—at least, in the way that I
well confirmed) is stated in such a manner that it define relevant.'^
can be subjected to empirical tests:
The reason for this requirement is set forth most
Sterling's Law: The depreciation of automotive clearly by Hempel when he defines "hage" as the
equipment is described by ap- product of man's height and age.'" As Hempel
plying the declining balance points out, the concept of hage has many qualities.
method at 1.5 the straight-line In his terms, it " . . . would have relatively high pre-
rate. cision and uniformity of u s a g e . . . " In accounting
You may not like my law since it deals with values terms hage would be "objective, verifiable and free
instead of costs. Please forget about that for a from bias." However, it lacks the quality of theo-
moment. Let us focus only on the structure of the retical import or relevance. Hage is irrelevant to
law: It has all the attributes of any scientific law, everything or, as Hempel says, " . . . we have no
the main one being that it is an empirically testable general laws connecting the hage of a person with
generalization. This means that, if an issue arose, other characteristics." Since it fails the test of theo-
we could resolve it. Suppose that an opponent of- retical import, no scientist bothers to calculate or
fered the following law: report hage.
Opponent's Law: The depreciation of automotive Hempel's example was fictional. If you want a
equipment is described by the real example, consider the concept of the cephalic
straight line method. index. At one time many scientists were busy cal-
culating and reporting cephalic indices in the belief
To choose between the two laws the only thing we that they were theoretically important. Brodbeck,
would have to do is make a sufficient number of using "useful" as a synonym for "theoretical im-
observations of exit values to decide whether port," writes:'^
Sterling's Law or Opponent's Law is correct.
Note the contrast. If depreciation is defined as a . . . the concept cephalic index can be quite ade-
conventional allocation of costs, there is no possi- quately defined in terms of the ratio of the width
bility of deciding between accelerated and straight of a person's head to its length, with the method
line—the issue will remain unresolved because it is of measurement precisely stated. Yet, for the
in principle unresolvable. If present depreciation is psychologist at least, a person's cephalic index is
defined to depend upon estimations of future lives an uninteresting fact. It is not interesting be-
and salvage values, then we can never know the cause a person's cephalic index has no connec-
present depreciation because we can never know tion, as far as we know, with any of his behavior.
the future. The proposed redefinition has the ad- There are, in other words, no laws connecting it
vantage of being a resolvable issue. with behavioral concepts. It is, therefore, not a
useful concept. Because cephalic index does not
Defining Concepts without Theoretical Import enter into laws, the concept is useless for pre-
Many of you will not like the law proposed in the dicting or explaining behavior. It lacks "mean-
previous section. You may say that my law is con- ing" in the sense of significance.
cerned with values and that accounting is con- Since the cephalic indices were irrelevant, since

FINANCIAL ANALYSTS JOURNAL / SEPTEMBER-OCTOBER 1975 D 3 3

they failed the test of theoretical import, scientists issues continues to increase. The purpose of this
quit calculating and reporting them. Thus irrele- article was to examine the reason for our inability
vance in science is sufficient reason for abandoning to resolve issues.
a concept. It seems to me that our inability to resolve issues
It may be that the concept of historical cost de- stems from the way in which we defme our prob-
preciation passes the test of theoretical import. lems. The definition of accounting problems pre-
However, I have not been able to discover any in- cludes resolution of those problems. Most ac-
stances where it is connected to other concepts or, counting issues are not "researchable" in the scien-
to put it another way, any decision models that tific sense of subjecting them to tests of empirical
specify the concept, despite a diligent search for and theoretical import. Instead, accounting issues
such laws or decision models. While doing the re- are defined so that they are matters of taste as in
search for Enterprise Income, I examined a num- art and the "research" consists of arguments
ber of economic decision models. In no case did whether one person's taste is better than another's.
the examined decision models specify such con- I conclude that a prerequisite for the resolution
cepts as "depreciation" (as we accountants define of accounting issues is the redefinition of those
it). Since such concepts were not specified, the in- issues. Specifically, the issues must be defined so
evitable conclusion is that they were irrelevant to that they are subject to scientific tests. Redefinition
those decision models. of the issues will not provide us with easy solutions
I wrote and talked about that conclusion for to our problems. On the contrary, there are many
some years. However, since it is impossible for me unresolved issues in science. There are disputes
to examine all laws or decision models, it is impos- among scientists just as there are disputes among
sible for me to prove that historical cost deprecia- accountants. The difference is that scientific issues
tion does not pass the test of relevance or theoreti- and disputes are subject to resolution by the appli-
cal import. That is, the fact that I have failed to cation of scientific tests. At the present time, ac-
prove its relevance does not mean that it is irrele- counting issues and disputes are defined so that
vant, since I may have overlooked something. For their resolution is in principle impossible. Resolu-
this reason, I published a challenge for someone to tion of the issues is a long way off, but the first step
prove that such concepts were relevant to one de- toward that resolution is to redefine the issues so
cision model."* Eight years have now elapsed since that they are testable. •
that challenge and no one has yet published a
reply. If accounting were a science, one of two Footnotes
things would have occurred: (1) someone would 1. The exchange among Cramer, Peterson and Scott
have rushed into print with a proof of the rele- concerns bad debts. Anthony and Defleise are re-
vance of such concepts or (2) the discipline would cent proponents of treating interest as a cost.
have quit calculating and reporting such concepts. 2. Accounting is the art of recording, classifying and
Again, consider the cephalic index. If someone summarizing in a significant manner and in terms of
could have demonstrated its relevance to just one money, transactions and events which are, in part at
law, scientists would have continued to calculate least, of a financial character, and interpreting the
and report those indices; in the absence of such a results thereof.
demonstration, scientists ceased calculating and If accounting were called a science, attention
would be directed (and perhaps limited) to the or-
reporting those indices. dered classifications used as the accountant's frame-
Until someone accepts my challenge, I will con- work, and to the known body of facts which in a
tinue to wonder about the relevance of such con- given case are fitted into this framework. See, Com-
cepts. I will wonder why we define such concepts mittee on Terminology, "Accounting Research and
as cost allocations instead of value measures. Since Terminology Bulletins, Number 1, Review and
historical cost depreciation is not empirically test- Resume," Accounting Research and Terminology
able and since, insofar as I know, it fails to pass the Bulletins, Final Edition (New York: AICPA,
test of theoretical import or relevance, why do we 1969), p. 9.
continue to define depreciation as a cost alloca- 3. Marvin L. Stone, "Public Confidence in Private En-
terprise—Let's Keep It," Journal of Accountancy
tion? That is, why do we defme it so that it is in (April 1968), p. 53. "But I must point out that ac-
principle unresolvable? counting principles are intellectual concepts. They
are not subject to the kinds of proof that are applied
Summary to laws of physical science. So legitimate differences
It is my contention that we do not resolve issues in of opinion within the profession itself are virtually
accounting. Instead we move from one unresolved inevitable."
issue to another, while the stock of unresolved See also, John Lawler, "The Quest for Ac-

34 D FINANCIAL ANALYSTS JOURNAL/ SEPTEMBER-OCTOBER 1975

 counting Philosophers," Empirical Research in among firms which use different methods of in-
Accounting: Selected Studies 1967 (Chicago: In- ventory pricing..."
stitute of Professional Accounting, 1968), p.88. See also, Gary M. Cadenhead, " 'Differences in
4. H. A. Finney and Herbert E. Miller, Principles of Circumstances': Fact or Fantasy?" Abacus (Sep-
Accounting Intermediate, 5th ed. (Englewood tember 1970). The author arrived at a similar con-
Cliffs, NJ.: Prentice-Hall, Inc., 1964), pp. 165-166. clusion.
"It. would be incorrect to suggest that the rules I have previously expressed doubt about the exis-
which establish whether a given accounting proce- tence of such circumstances: "The disputes are em-
dure is acceptable or unacceptable are in the nature pirically unresolvable since there are no observa-
of principles like those found in physics or tions which could be used to confirm or disconfirm
chemistry. Accounting principles are more properly either convention. It is sometimes said that the 'ap-
associated with such terms as concepts, conventions
propriate' inventory or depreciation method 'de-
and standards. It is important to remember that ac-
pends upon the circumstances' but I am unable to
counting principles are mari-made, in contrast to
discover any particular circumstances' which imply
natural law.
any particular method. My inability to discover the
Principles cannot be established in accounting, as circumstances does not prove that they do not exist
they are in the realm of natural sciences, by experi- but the fact that no one has yet made them explicit,
mentation; nor have they been determined, as in the even though the problem is a pressing one, casts
law, by authoritative pronouncement,..." some doubt upon their existence." See, Sterling,
5. Peter Caws, The Philosophy of Science (Pnnceton, "An Explication and Analysis of the Structure of
NJ.: D. Van Nostrand Co., Inc., 1966), pp. 85-86. Accounting Part Two," Abacus (December 1971),
The author draws the distinction: "The same state- p. 160n.
ment can at different times and under different cir- 12. See, Arthur L. Thomas, The Allocation Problem
cumstances be a principle, n definition, or a law. If in Financial Accounting T/ieory, Studies in Ac-
its truth is asserted and not considered liable to em- counting Research # 3 (Menashaw, Wisconsin:
pirical challenge, it is a definition; if not—that is, if American Accounting Association, 1969), for an
it is taken to be empirically significant—then it is a extended analysis of the unresolvability of alloca-
principle if it contains theoretical terms, otherwise a tion problems.
law. A principle turns into a law if what before See also. Sterling, The Theory of the Measure-
could not be observed becomes observable by virtue ment of Enterprise Income (Lawrence: University
of some advance in experimental technique." Press of Kansas, 1970), Ch. V-VI, for a discussion
See also, Karl R. Popper, The Logic of Scienti- of measurement.
fic Discovery (New York: Science Editions, Inc., See also. Sterling, "The Going Concern: An
1961), pp. 40-42, for an extensive discussion of the Examination," Accounting Review (July 1968),
falsifiability requirement in science. pp. 479-499. The author discusses the confusion of
6. Lawler, p. 88. temporal locations in accounting.
7. Warren S. Torgerson, Theory and Methods of 13. Committee on Terminology, p. 24.
Scaling (New York: John Wiley & Sons, Inc., 14. Sterling, Enterprise Income, p. vii.
1967), p. 1. 15. Sterling, Enterprise Income, p. 50. "The theory
8. Joseph R. Daughen and Peter Binzen, The Wreck concerning the solution of the problem specifies the
of the Penn Central (Boston: Little Brown and relevant information."
Co., 1971), p. 224. If a property is specified by a decision model,
9. C. H. Firth, Oliver Cromwell and the Rule of the then a measure of that property is relevant (to that
Puritans in England (London: Putnam, 1938), p decision model). If a property is not specified by a
453. decision model, then a measure of that property is
10. See Robert R. Sterling, "On Theory Construction irrelevant (to that decision model). See, Sterling,
and Verification," Accounting Review (July 1970), "Decision Oriented Financial Accounting," /I c-
p. 446, for a brief discussion of the distinction be- counting and Business Research (Summer 1972),
tween the empirical and nonempirical sciences. For P- 199.
a more extended discussion see Caws or Popper. 16. Carl G. Hempel, Fundamentals of Concept For-
11. In addition, consider the fact that the circumstances mation in Empirical Science, International Encyc-
upon which the selection is supposed to be based lopedia of Unified Science, Vol. I & II (Chicago:
have never been specified. See, Lanny G. Chasteen, University of Chicago Press, 1952), p. 46.
"An Empirical Study of Differences in Economic 17. May Brodbeck, ed.. Readings in the Philosophy of
Circumstances as a Justification for Alternative In- the Social Sciences (New York: The MacMillan
ventory Pricing Methods," Accounting Review Co., 1968), p. 8.
(July 1971), p. 508. The author attempted to dis- 18. What is historical cost relevant to?
cover the circumstances which correspond to dif- Since this is a [xilemic, I will state the proposi-
ferent inventory methods being applied in practice. tion positively so that the reader can demolish it:
He concluded, "Generally, no significant differences (1) "Relevance is the primary standard..."
in economic circumstances were consistently found and " . . .all information that does not meet the

FINANCIAL ANALYSTS JOURNAL / SEPTEMBER-OCTOBER 1975 D 3 5

 standards should be excluded." phers." Empirical Research in Accounting: Select-
(2) Historical cost is irrelevant to all economic ed Studies 1967. Chicago: Institute of Professional
decisions (Sterling's allegation). Accounting, 1968, pp. 86-92.
(3) Ergo, historical cost should be excluded. Petersen, Russell J. "Incompatibility of Bad Debt 'Ex-
The quarrel will probably be with (2) and, again pense' with Contemporary Accounting Theory: A
being deliberately polemical, 1 challenge any one Comment." Accounting Review (October, 1973), pp.
to demonstrate that historical costs are relevant to 777-778.
economic decisions. Since (2) includes a universal Popper, Karl R. The Logic of Scientific Discovery.
quantifier it will take only one counter case to New York: Science Editions, Inc., 1961.
falsify it. See, Sterling, "ASOBAT Review," pp. Scott, Edward R. "Bad Debts: Take Two: A Comment."
110-111. Accounting Review (October 1973), pp. 775-776.
Sterling, Robert R. "An Explication and Analysis of the
References Structure of Accounting Part Two." A bacus (Decem-
Anthony, Robert. "Accounting for the Cost of Equity." ber 1972), pp. 145-162.
Harvard Business Review (November-December "Decision Oriented Financial Accounting."
1973), pp. 88-102. Accounting and Business Research (Summer 1972),
Brodbeck, May, ed. Readings in the Fhiiosophy of the pp. 198-208.
Social Sciences. New York: The Macmillan Co., _. "The Going Concern: An Examination."
1968. Accounting Review (July 1968), pp. 481-502.
Cadenhead, Gary M. " 'Differences in Circumstances': _. "On Theory Construction and Verifica-
Fact or Fantasy?" Abacus (September 1970), pp. 71- tion." Accounting Review (July 1970), pp. 444-457.
80. _. "A Statement of Basic Accounting Theory:
Caws, Peter. The Fhiiosophy of Science. Princeton, N. A Review Article." Journal of Accounting Research
J.: D. Van Nostrand Co., Inc., 1966. (Spring 1967), pp. 95-112.
Chasteen, Lanny G. "An Empirical Study of Differences The Theory of the Measurement of En-
in Economic Circumstances as a Justification for Al- terprise Income. Lawrence: University Press of Kan-
ternative Inventory Pricing Methods." Accounting sas, 1970.
Review (July 1971), pp. 504-508. Stone, Marvin L. "Public Confidence in Private Enter-
Committee on Terminology. "Accounting Terminology prise—Let's Keep It." Journal of Accountancy
Bulletins, Number 1, Review and Resume." Ac- (April 1968), pp. 52-57.
counting Research and Terminology Bulletins, Thomas, Arthur L. The Allocation Froblem in Finan-
Final Edition. New York: AlCPA, 1969, pp. 1-32. cial Accounting Theory. Studies in Accounting
Cramer, Joe J., Jr. "Bad Debt 'Expense': Not a Member Research # 3 . Menashaw, Wisconsin: American Ac-
of the Class of Data for Measuring Operating Income: counting Association, 1969.
A Reply." Accounting Review (October 1973), pp. Torgerson, Warren S. Theory and Methods of Scaling.
779-784. New York: John Wiley & Sons, Inc., 1967.
"Incompatibility of Bad Debt 'Expense'
with Contemporary Accounting Theory." Accounting
Review (July 1972), pp. 596-598.
Daughen, Joseph R. and Peter Binzen. The Wreck of QUARTERLY DIVIDENDS SINCE 1935
the Fenn Central. Boston: Little, Brown and Co.,
1971. NATIONAL
DeFleise, Philip L. "The Economic Case against Capita-
lization of Leases—A New Look at Facility Ac- DISTILLERS
counting." Institutional Issues in Fublic Ac-
counting. Edited by Robert R. Sterling. Scholars sCHEMICAL
Book Co., 1974.
Finney, H. A. and Herbert E. Miller. Frinciples of Ac-
CORPORAnON
DIVIDEND NOTICE
counting Intermediate. 5th ed. Englewood Cliffs, N. The Board of Directors has declared
J.: Prentice-Hall, Inc., 1964. a quarterly dividend of 3O0 per share
on the outstanding Common Stock,
Firth, C. H. Oliver Cromwell and the Rule of the payable on September 2. 1975 to
stockholders of record on August 11.
Furitans in England. Black and White Library 1975. The transfer books will not close.
Edition. London: Putnam, 1938. July 24, 1975
Hempel, Carl. G. Fundamentals of Concept Forma- RAMSEY E. JOSLIN,
Vice President-Financial
tion in Empirical Science. International Encyclope- DIV1SI0NS:U& INDUSTRUL CHEMICALS C a
dia of Unified Science, Vo. I & II. Chicago: Univer- NATIONAL DISTILLERS PRODUCTS CO.
BRIDGEPORT BRASS CO.
sity of Chicago Press, 1952. ALMADEN VINEYARDS, INC
TEXTILE
Lawler, John. "The Ouest for Accounting Philoso-

36 a FINANCIAL ANALYSTS JOURNAL/ SEPTEMBER-OCTOBER 1975