Philosophy of language and accounting

2.1 Philosophy of language

Cultural diversity creates differences in the understanding of concepts, leading to diversity in the definitions of common notions. Searle’s and Wittgenstein’s philosophies of language provide nuanced accounts of precisely how the meaning of accounting concepts or principles can vary according to cultural and historical context. Two ideas are especially critical for understanding the impact and meaning of the relevant principle. First, Searle’s concept of social facts as collective intentionality, and second Wittgenstein’s emphasis that the meaning of words and ideas varies not only according to the different cultural meaning of the words themselves, but also according to the specific social practices in which the language usage is set.

The philosopher Searle (1995) explained, based on the concept of “intentionality,” how social objects come into being. In this regard, it is possible to distinguish between subjective dependent facts (i.e. dependent on individual intentionality) and social facts (dependent on collective intentionality). Searle (2006) defined a social fact as “any fact involving collective intentionality of two or more human or animal agents” (p. 16). According to Searle, people have the ability to share beliefs and desires, which – in certain conditions – can give rise to a specific type of social fact, namely, “institutional facts.”

We broadly accept the idea developed in a number of recent papers (e.g. Shapiro, 1997, 1998; Alexander and Archer, 2003; Mouck, 2004; Alexander and Ionascu, 2008), not always quite consistent with each other, applying to the accounting domain the concept of institutional and inter-subjective reality developed by Searle (1995, 2006, 2009). As Alexander and Archer (2003, p. 5) argued, “By virtue of collective intentionality, ownership claims, income, and other conceptual objects of accounting can, under appropriate conditions, be institutional facts.” Following this thinking, institutional facts, although ontologically subjective (as they require human practices to sustain their existence), are epistemologically objective, meaning they have an effect that is agreed upon. That is because, being inter-subjectively constructed by means of collective intentionality, institutional facts become objectified; in other words, they are not dependent on a particular human being’s attitude toward them.

Consider the statement “Inventory €100 in accordance with EU-endorsed IFRS.” Social constructions are legion: money, euro, cost, EU, inventory and IFRS. Physically identical buildings can be recorded, in appropriate circumstances, at “cost” (IAS 16 para. 30), at “fair value” (IAS 16 para. 31 or IAS 40), or at “the lower of cost and net realisable value” (IAS 2). So in what sense is a fair presentation “real”? Different “realities” are involved. A fair presentation must communicate the substance of the relevant inter-subjectively constructed “reality.” This is impossible without collective agreement, collective intentionality, about that particular inter-subjectively constructed “reality” within the community of actors involved. And of course no collective intentionality is possible without effective communication and mutual understanding within and across the collective.

Based on this concept of social facts, both animals hunting together and the European Union Parliament issuing a Directive are examples of social facts, involving collective intentionality. Institutional facts are the result of the collective attribution by human beings of status functions to parts of reality. Indeed, human beings have the capacity to ascribe functions to objects only by virtue of the collective assignment (Searle, 2006). For example, a yellow line on a floor can have the function of a barrier, not by virtue of its physical features, but because of the collective assignment to that line of the status of boundary marker. As Searle (2006) asserted, “status functions are the vehicles of power in society.” Human beings, by accepting the status functions, create obligations, responsibilities, duties, rights, etc., to which Searle refers as deontic powers. With “collective intentionality,” therefore, Searle (1990) maintained that human beings are able to participate in cooperative behaviors, adopting complex devices such as language or other tools based on symbolization to achieve collaborative actions and beliefs.

A necessary condition for institutional facts and internal reality is concerned with the communication of concepts between human beings – with words. Perceptions of economic reality are personal and subjective. How, if at all, can such perceptions be explained and communicated from one human being to another? According to Saussure (1973), language is a system of signs. A sign involves two elements: first, that which signifies, called the signifiant or signifier, e.g. a word; second, the idea signified, called the signifié or signified. Saussure argued that the sign is arbitrary in all of three respects: sign, signified and the connection between.

Searle (1995, Chapter 3; 2009) argued for an unavoidable relationship between institutional facts through collective representation and language. Beyond the obvious similarities regarding shared beliefs (institutional facts) and shared meanings (language) is the more fundamental point that the move from object (e.g. pebble) to status function by means of collective representation (e.g. money) is possible only if there is a mechanism for creating, and communicating, this collective representation. It is language that gives humans the capacity to represent.

So can we postulate a collective representation, in this technical sense, about SoF? We now bring in Wittgenstein who elucidated an additional perspective on language. The most fundamental difference between Wittgenstein and Searle is that Wittgenstein even questioned whether humans can actively and without presumptions collectively create sign, signified and the connection between both (e.g. Baker and Hacker, 2009). For Wittgenstein the sign provided by language is not independent from the signified reality, but it is intrinsically and inseparably linked to certain forms of life.

Wittgenstein understood concepts such as SoF in his later philosophy as receiving their meaning from the way they are used (Wittgenstein, 2005, Sections 20, 43, 421; transposed to accounting by Lyas, 1993; Dennis, 2008). Wittgenstein chose this quite uncommon starting point because he considered it to be a misleading simplification to reduce the link between language and reality to one of an arbitrarily chosen and nominalist correspondence between concepts and reality (see Kenny, 1974; Baker and Hacker, 2009). This has several important consequences.

A first consequence is that one can only understand the meaning of concepts when focusing on the contexts, the language game, in which certain concepts play a certain role and the relationships through which those concepts receive their meaning (e.g. Wittgenstein, 2005, Sections 109, 471, 583). He pointed out, for example, that a stick becomes a lever only when it is used for this purpose (Wittgenstein, 2005, Section 6). This aspect expands on Searle, as it highlights that beyond the yellow line being a social fact, the process of being acknowledged as having a specific meaning is ensured by a living practice in which a particular concept plays a role. Wittgenstein denied arbitrariness because he pointed out that communication with the help of language presupposes that typical or entrenched relationships or sets of contexts/practices exist. He noted that any concept has a meaning within such a language game only because its sense is identical with or – better – is its role in the particular language game (Wittgenstein, 2005, Section 49).

A second important consequence is that the meaning of a given concept is not inherently determined. Indeed, it has no intrinsic essence beyond its use (e.g. von Savigny, 1969; Wittgenstein, 2005, Section 432). Therefore, one needs to observe what happens with a concept in different contexts rather than assuming that a concept has a fixed and stable essence (Pitcher, 1967). Its degree of vagueness, or complexity, its stability over time, its definition, all this might change with the differing contexts in which the concept is applied in practice (see Pitcher, 1967; Stegmüller, 1989; Wittgenstein, 2005, Sections 47, 79, 87, 88, 593). Still such a concept might in all of these circumstances be perfectly serviceable (Dennis, 2008).

Despite their very different starting points, strong similarities emerge between Searle’s “collective assignment” and Wittgenstein’s “rules of the game.”

In our context, the question also emerges what happens with concepts that are applied in other contexts than within their natural language games. One thing is clear for Wittgenstein: a concept becomes nonsense when applied outside of its language game or in the wrong language game (Wittgenstein, 2005, Sections 39, 40). However, if in given operating language games similarities exist, a concept from another language game can be compared and it assumes a meaning and it can be introduced within that language game, however, slightly changing its meaning to fit to this new context. According to Wittgenstein (2005, Section 140), understanding is nothing but being able to check whether a certain concept can be applied to a certain situation (see also Wittgenstein, 2005, Sections 145, 206, 232, 233). Explaining the meaning of a concept then amounts to providing different examples of its use (Wittgenstein, 2005, Sections 71, 72). Remembering the meaning of words, on the other hand, is nothing but remembering how this given word was applied in the past (Wittgenstein, 2005, Section 379).

Coming back to our question about an SoF collective representation, following Wittgenstein and consistent with Searle, the most likely finding can be described as follows. We can postulate a series of collective representations of SoF, each understood, despite fuzzy boundaries, by different related, but non-identical, examples of use in varying “collective representations” (Searle) or varying “language games” (Wittgenstein), both these terms being used in a precise technical sense. We explore the validity of this proposition in the remainder of the paper.

2.2 Translation issues in accounting

Translation aims at functional equivalence, such that the reader of a translation should read an exact translation of the meaning instead of a word-by-word translation. This is, of course, standard translation procedure, also applied by the Translation Centre for the Bodies of the European Union[1]. The language of first drafting EU Directives is now most likely to be one of the three official “working languages,” English, French or German. Member States of all other 23 languages base their implementation on translations, which are supposed to convey the same content. This is where a problem may arise.

Relying on the Sapir–Whorf hypothesis, according to which speakers of different languages perceive the world differently, as they are influenced by their own language and culture, Evans (2004) and Evans et al. (2015) pointed to the fact that accounting regulations include culture-dependent concepts, like the true-and-fair-view principle. The consequence is that accounting and legal terms are understood differently in locations with different language and culture. Specific burdens in the functional translation of accounting concepts include the differences between the cultural and historical context of common law and civil law countries, or between Western and former socialist countries of Eastern Europe.

Along this line, although translators can transfer a specific term from one language to the other, readers of the translated text will not decode it in exactly the same manner as native-language speakers reading the original text. Individual culture, history and habits, used as decoder support, differ. “Equivalence” relates to a variety of possible “forms of correspondence between the source text […] and the target text […]” (Kettunen, 2017, p. 41; Pym, 2014). Ignoring the technical linguistics-related details here, earlier theories proposed that a “natural equivalence” should normally exist, and translation is the process of finding it. Our theoretical framework, however, might suggest that any expectation of such unproblematic natural equivalence between languages, and therefore “automatically valid” translation, is likely to be problematic.

Transferring these considerations to the present discussion leads to the expectation that the meaning of the SoF notion depends on the language game (broadly the country) in which it is applied, unless there is a general meaning to cover all countries, i.e. with collective assignment across countries. This leads us to expect an increase in doubts in different contexts, where the notion is new or applied differently.

To summarize, we can suggest that perceptions of, interpretations of and chosen applications of a principle such as SoF can only be agreed (harmonized) as social constructs, ever-changing by definition, through collective intentionality within groups of like-minded thinkers. A necessary condition for the creation of “like-minded thinking” is effective communication of “the thoughts,” through the highly imperfect medium of language – itself an ever-changing social construct.

2.3 Methodology

We focus our analysis on the SoF principle for several reasons. First, while new in some countries, it expanded worldwide through the use of IFRS[2]. Second, the principle receives special treatment in the Accounting Directive, where Member States may opt-out. Third, the principle is undefined at the European level. We choose our countries to reflect the diversity in Europe and its impact on local regulations and the difficulties in enacting common Directives.

We identify four different country-groups, which already suggest diverse “collective intentionalities” – UK common law, France and Italy civil code, Germany and Austria tax code, and Poland and Romania ex-communist. Our investigation per country includes the presence (or absence) of the SoF principle (or a similar concept) in the respective accounting history, and before the Accounting Directive, and its integration during the implementation of the Directive. Different approaches to implementation, and different wording may result from (and in) different concepts and uses.

3.1 In the Accounting Directive

The final version of the 2013 Accounting Directive, taking the English wording, seems to establish the notion of SoF, albeit in a very vague way:

Article 6.1(h): “items in the profit and loss account and balance sheet shall be accounted for and presented having regard to the substance of the transaction or arrangement concerned;”

Article 6.3: “Member States may exempt undertakings from the requirements of point (h) of paragraph 1.”

The word “substance” is not defined. Further, the requirement is simply to “hav[e] regard to” the substance. “Having regard to” means that one must look at it and genuinely consider it in the overall context concerned, but one may then go against it if it is rational within that (socially constructed) context so to do. No explicit obligation to follow the undefined substance is implied. Therefore, at first glance, Article 6.1(h) implies only a vague priority of an undefined “substance,” presumably over an unmentioned, possibly legal, form.

However, the Directive then negates the requirement completely at the Member State level, as under Article 6.3 the States “may exempt undertakings” from the requirement, with no alternative being given. In other words, whether – and, if so, to what extent – to implement the “having regard to the substance” requirement is completely left to the Member States. Member States may choose to base their accounting laws on legal form only, but may also exempt only specific undertakings (e.g. small firms, specific industries). They may even choose different principles for individual and consolidated statements, as preliminary recital 35 allows[3]. It seems that Member States were not able to agree on harmonization of the principle (or its rejection) and had to compromise by allowing Member State discretion (Florstedt et al., 2015). This hinders harmonization, consistency and comparability of statements among firms and countries.

While the analysis so far has referred only to the English version, the overall picture is much more complex. Details are discussed in the country sub-sections below. To highlight the key result, three of the language versions, English, French and Italian, appear to say similar things. But in the other four countries, Austria, Germany, Poland and Romania, different translations introduce the concept of “economic” (wirtschaftlichen, economic, ekonomicznej) in relation to the “substance.” Surprisingly, the words “form” or “legal form” do not appear – express or implied – in any of the versions.

3.2 In seven European countries

3.3 Synthesis

From the country analyses, several issues emerge. The first observation is the divergence in the implementation of the Directive. A synopsis in English (authors’ translations) of the different versions of the SoF principle in local enactments reveals the differences as shown in Table I.

While the different versions of the SoF principle back-translated into English are taken out of context and may suffer from the individuality, and possible idiosyncrasies, of the different authors, still they clearly show that wordings differ. Instead of “substance,” “business significance,” “economic substance” and “economic substance or nature” are used; instead of “form,” “legal form” and “appearance” are used, or no reference is made to any kind of “form” at all. The UK speaks of “transactions and other events and conditions,” whereas the Accounting Directive uses “transaction or arrangement,” Poland refers to “events, including business transactions,” etc. This after the different language versions of the Directive itself already presented differences in wording.

All countries have some way of including an SoF requirement in their accounting principles, explicitly or implicitly. No country within our survey claimed to opt out.

Generally, the local enactments attempt to follow the local translations used for the Directive, unless a close concept was already present under local rules and was not changed, due to local tradition or previous implementation. Italy has quite deliberately changed the wording of its enactment from the previous “economic substance,” to “substance,” exactly following the wording of the Directive in Italian. Romania has maintained its previous usage of “economic” substance (derived from earlier incorporation of IFRS terminology), with the explicit support of the Romanian version of the Directive. Poland also maintains the general thrust followed since 2000, of “business significance,” with a variety of different and individualistic changes in terminology, but not following the wording of the Directive in its Polish version – deliberately changing “having regard to” to the much more definitive “according to,” for example. The UK has very much ignored the detailed wording of the Directive, deliberately – almost provocatively – bringing in “legal form” in order to contemptuously dismiss it.

Germany refrained from an explicit introduction of the concept into its accounting laws under the claim that the national regulations already fully contained the requirement. German practice – and its courts – indeed follows an economic substance concept. Yet, it is one entirely constructed as an expression of the teleological method of civil law – in this case historically mostly focused on the needs of tax law and not on the needs of investors or other stakeholders as far as these go beyond principles that are already part of the legal system. France seems more obscure, perhaps vague, in its enactment. Prior wordings are very much retained, and the (French) wording of the Directive seems deliberately and totally ignored. Finally, there is no overall substance principle, which the Directive proposes, but former sincerity (sincérité) remains, for which the Directive gave no requirement at all. There is, however, a case-by-case “substance-based” treatment of a number of specific types of transaction, although with contradicting rules.

When looking for common features among the respective concepts, it appears that France and Germany both consider an explicit implementation of any kind of SoF at this time as unnecessary, and both would claim to have fully enacted the contents of the Directive. Poland and Romania both have an accounting history which is rooted in statistical requirements of the communist economic regime, where there was no room for any notion of SoF. Subsequently, both countries readily implemented the principle with a view to IFRS. Germany and Austria both have a strong tax background in their respective SoF principles, and a tradition of non-codification; the almost literal implementation of the Accounting Directive in Austria, however, is in contrast to the non-explicit implementation in Germany. This suggests consistency with the proposition of overlapping/conflicting social realities/language games as developed in Section 2.

Considerable differences in accounting for leases or own shares, for instance, illustrate that in areas where accounting for the substance rather than the form is frequently required, still no harmonization exists for countries. Finance leases are expensed in France and Italy; own shares are shown in assets in France and Poland. Despite a common principle generally enacted – even if not exactly named the same way – the consequent booking rules are different.

Within our sample of seven countries we demonstrate that, in spite of all applying the principle, there is an astonishing degree of diversity in the reactions to the wording of the Directive (Table II). The social construct of SoF may be agreed through collective intentionality within groups of like-minded thinkers, but these groups are distinct per country, and not uniform across Europe. There is little consensus, and less “harmonization.”

4.1 Only a question of translation?

Our sample of countries is definitely diverse: seven countries, representing both common law and civil law systems; two former communist countries; and five different translations of the English version of the Directive. A relevant illustration of the cultural influence in translation is the need of some civil law regulators to define the term substance more precisely, either by adding the adjective economic (Germany, Austria and Romania), or by using a different term (business significance in Polish). This is a functional translation, which helps the reader understand the meaning of the translated concept, by connecting it to a common correlation, delivered by the cultural background of the three countries: economic vs legal in Romania and Poland, and economic naturally being determined legally in Germany.

Using the philosophical lenses, we actually question the common understanding of the SoF principle. The use of different wordings, especially when these wordings existed before the Directive, the lack of any definition at the European level and the resulting variety in practical rules, all tend to confirm that different communities within Europe share different frames of social and institutional facts. They are, however, overlapping between countries, such as the tax background of the principle in Austria and Germany; the communist past in Poland and Romania; civil law in Austria, France, Germany and Italy. There is no common European (accounting) community yet, but there is a variety of differently overlapping sub-communities, which goes across different countries. This situation, and not merely bad translation, represents a significant obstacle to harmonization.

4.2 Where is the principle defined?

The origins of the international and British SoF principle (meaning the idea rather than the precise wording) appear to lie in practice, in procedures adopted and applied in concrete cases without explicit argumentations (Rutherford, 1988), which could explain the undefined meaning of the principle today.

Despite its technical aspects, accounting remains about social facts. Notwithstanding a now more than 60-year construction, Europe has not managed to build a common (accounting) language game. One reason could be that in the Directive, the main missing point is any definition of the concept itself. Without any common description of (economic) substance and/or (legal) form, each country had various options: craft its own definition, regulate the principle without official definition or use a supposed close concept taken from local roots. The result could only be diversity. Perhaps, however, the omission of any “common description” simply recognizes the permanent inevitability of the implications of our theoretical underpinnings. The social constructions, across both country communities and across user group communities, are perhaps too multitudinous, ever-changing and too diverse to allow any other result.

4.3 Why should substance and form differ?

The objective of SoF is of course to obtain a TFV, in all the countries in the sample, although given the subjectivity and institutional specificity of both these concepts this is little more than a legalistic truism. Certainly the conclusions as regards the rules quite differ.

The TFV induces a potential override of any rule that would prevent it, at least in the UK, France and Romania. In Italy, however, the override is limited to exceptional cases, suggesting that the rules should normally be “fair” enough to be followed. The case reveals that the principle comes with an objective to inform the reader on an economic patrimony, instead of a pure property rights approach, but this is introduced in a system where codified norms and the legal classification of transactions and contracts prevail.

In France, some authors supported a substance-based analysis of law (Raybaud-Turrillo, 1995), noting that the Civil Code already pointed at substance. Like Romania, France considers that the legal form should not disguise the substance. The accounting principle could even be redundant: the question is whether to include it formally in accounting rules, as redundancy may lead to interpretation issues (a view similar to the German one).

In Germany, the “economic view” suggests that no rule could be valid if it goes against this principle. A civil law notion, the potential basis of a transaction between parties, may have economic consequences determined under law. The figure arising from a particular legal form can indeed be changed to an equally legally determined substance figure, but only in accordance with legally established practice, not by the individuals concerned. The only overriding option in Germany is limited to the legal sphere: the interpretation of transactions made available by law is supposed to be fair in the end. In Austria, by contrast, the economic view does override the mere substance for tax law as well as for accounting purposes. The Austrian economic view therefore goes beyond application of the teleological interpretation of the law. The different notion of economic view in Germany and Austria, in spite of common roots at the beginning of the last century, makes another case for the social construction of legal realities.

4.4 Who defines the substance?

We show that similar concepts following close objectives ultimately have opposite consequences on the authority of accounting rules. The question is to understand who should define and reveal this substance to the users of financial information.

Under common law systems, rules are derived from the recognition of practices, observed in the field. The civil law system names, defines and classifies the notions, to give individuals a framework to interpret their own business. Germany and Austria have maintained their traditional approach. Romania and Poland underscore the importance that differences between substance and form be kept exceptional, with Romania asking the preparer of contracts to align form to substance. This is perhaps the easiest way to implement SoF in a system where the traditional “collective representation/language game” was very different from that inherent in SoF. Since accountants are used to following the legal form of a document, then it is necessary to avoid any conflict between the legal form and the economic substance. Besides, in each country the “legal form” is likely to carry with it unavoidable rights and obligations, often unstated but automatically embraced. Italy modified the Civil Code to include the notion, while France did not, as the substance-based intention of parties should already be formal. Both attempt to define the accounting consequences of a civil law concept, having one accounting rule per transaction, but they fail in some cases that should be reviewed, or distinguished. In this sense, those six countries show a civil law way to integrate this principle, while the UK only acknowledges a concept already taken for granted. We find here the traditional opposition between civil law environments, where the telos is under central hands (law, tax and judge), and common law systems, where the substance should be revealed by either party in a contract.

4.5 The question of users

Finally, the situation may also reveal different uses throughout Europe and even inside each country, of accounting information. In our sample of countries, it is not clear that accounting information has been given the same collective assignment, according to Searle’s concept. Accounting information can be used in tax calculation, in some legal procedures (dividend distribution, litigation and liquidation) and also to inform various users on the performance and financial situation of the firm. Contrary to IFRS standards, local rules in Europe may not formalize the role of accounting information, especially for individual accounts. If the role of information differs significantly, then the subjectively determined “substance” behind that information can logically be expected to differ too.