Future tense and psychological distance

When a verb refers to the future, some languages require explicit marking of that fact. A recent paper presents evidence that companies in countries using those languages are slow in reporting that their goodwill has lost value. The paper suggests that this is because speakers of those languages perceive the future as psychologically more distant than speakers of other languages.

The paper is The Effect of Time Orientation in Languages on the Recognition of Goodwill Impairment Losses, by Ahmad Alshehabi, Hussein Halabi, Godfred Afrifa, in The International Journal of Accounting (2023), online ready at https://doi.org/10.1142/S1094406023500117

In this post, I look at:

  • how languages mark verbs as future.
  • the evidence the authors found.

The original longer version of this post appeared on my accounting blog Accounting Miscellany at Does future tense help keep goodwill alive? – Accounting Miscellany
This condensed version focuses only on linguistic issues. The full version also covers accounting issues.

The authors’ list of references includes useful references on 4 topics, among others: language; culture; discounting effects of future events; and accounting. To provide a resource for future work, I have copied their references on those 4 topics into the appendix to the full version of the post on Accounting Miscellany.

How languages mark verbs as future

The authors adopt a classification of languages—made originally by Chen (2013)—into 2 groups:

  • Strong-FTR languages’, such as French or English. These languages mark verbs explicitly to show when the verb refers to an event in the future. For example, French marks verbs as future by attaching an inflectional ending to a future stem: sera (‘will be’) versus est (‘is’). Here, ser- is the future stem of the verb être (‘be’) and –a is the 3rd person singular future suffix; est is the verb’s 3rd person singular present form. And English marks verbs as future by inserting before the infinitive a modal verb or auxiliary verb: for instance, will be or is going to be, contrasting with present tense is.
  • Weak-FTR languages’. Some weak-FTR languages (eg Mandarin and Finnish) have no means of marking a verb as future.  Other weak-FTR languages have such means but often do not use them (eg German).

The acronym FTR stands for future time reference.

Weak-FTR languages can still mark the future

Even though weak-FTR languages do not mark verbs as referring to the future, speakers of those languages can distinguish the future from the present—for example, by including adverbials such as ‘next year’. But those languages do not compel speakers to decide, every time they speak, whether a verb is to be marked as future.

The distinction between weak-FTR and strong-FTR languages illustrates a well-known comment by the linguist Roman Jakobson:

‘Languages differ essentially in what they must convey and not in what they may convey’.

Which languages are in which category?

For the countries tested in Alshehabi, Halabi and Afrifa (2023), Table 1 shows whether the language(s) spoken there is (are) strong-FTR or weak-FTR.

Strong-FTRWeak-FTR
AustraliaAustria
FranceBelgium
GreeceDenmark
IrelandFinland
IsraelGermany
ItalyHong Kong
New ZealandNetherlands
PolandNorway
PortugalSweden
South Africa 
Spain 
United Kingdom 
Table 1.  Countries tested by Alshehabi, Halabi and Afrifa (2023).

In Table 5 of an online appendix Chen (2013) lists many more languages, classifying them all as strong-FTR or weak-FTR. Interestingly, that table 5 classifies European Portuguese as strong-FTR but Brazilian Portuguese as weak-FTR. https://assets.aeaweb.org/asset-server/articles-attachments/aer/data/april2013/20111628_app.pdf

Binary split or continuum?

Most of the authors’ analysis uses Chen’s binary split between weak-FTR and strong-FTR languages. In reality, languages sit on a spectrum, with no clear boundary between the 2 classes. So, the authors also carried out tests using 2 continuous measures of FTR developed by Chen (2013). 

Those measures were a verb ratio and a sentence ratio. These ratios show how often grammatical marking of the future occurred in weather forecasts retrieved from the internet. The authors report that these tests produced results similar to the results of their main tests using the binary split.

The use of weather forecasts as texts used in developing the 2 continuous measures of FTR is interesting. Weather forecasts are one context where some weak-FTR languages do not mark verbs as future—even if they mark verbs as future in some other contexts.

Chen (2013) defines weak-FTR languages as languages not requiring the obligatory use of grammaticalized future-time reference in (main clause) prediction-based contexts. In this definition, a prediction is a statement about the future with no intentional component: a canonical example is predicting the weather. Chen took his definition from Dahl (2000), though Dahl labelled that category as ‘futureless’ languages.

Previous research on future time reference

The authors say that previous research suggests speakers of weak-FTR languages are more likely (than speakers of strong-FTR languages) to behave in a way oriented toward the future. They:

  • are more physically active, smoke less and are less likely to be medically obese.
  • save more, and hold more cash as a precaution.
  • are more future oriented, and are more willing to accept short-term costs in return for long-term rewards. For example, they invest more in research and development, protecting the environment, sustainability and social responsibility.
  • avoid decisions with negative future consequences—because these consequences seem more imminent.
  • are less likely to engage in tax avoidance.

Alshehabi, Halabi and Afrifa (2023) say there has been little previous research on how language affects corporate financial reporting (accounting). They report that:

  • Kim et al (2017) concluded that speakers of Strong-FTR languages are more likely to manage reported corporate earnings.
  • Gotti et al (2021) dispute that conclusion. They argue that the findings reported by Kim et al (2017) disappeared after further analysis that looked not just at individual languages but also at language families (eg Indo-European).

How does the psychological mechanism work?

This paper, like some of the previous research cited, finds some evidence of an association between some observable behaviour and whether a speaker’s language routinely requires the speaker to explicitly mark verbs that refer to future events. Those papers speculate that the future is psychologically more distant for speakers of strong-FTR languages than for speakers of weak-FTR languages.

I haven’t read all the underlying papers, but I have seen nothing confirming that apparently speculative explanation:

  • I agree that having to mark each verb explicitly as future (or not) might make speakers more conscious of the distinction between future and present.
  • On the other hand, having to mark that distinction every single time might just as well make the distinction less salient, through constant repetition (just as we learn to tune out background noise).

I have no information to tell me which of those plausible mechanisms is more likely to operate.    

The evidence found by the authors

The authors found evidence that companies in countries with weak-FTR languages:

  • were more likely to report goodwill impairment losses than in countries with strong-FTR languages.
  • reported larger goodwill impairment losses.
  • were more likely to report goodwill impairment losses whose amount was ‘normal’ or ‘expected’. The measures of ‘normal’ and ‘expected’ losses were functions only of economic and accounting variables selected by the authors.

Background information: goodwill and impairment losses
When a company buys another company, the acquiring company brings into its own balance sheet all the assets of the acquired company, as well as an asset called ‘goodwill’. The value of goodwill is the difference between the value of the acquired company’s whole business and the total value of all its assets.

Each year after the acquisition, the acquiring company checks whether the goodwill has kept its value. If the goodwill has lost value, the acquiring company reports an expense, known as an ‘impairment loss’.  

The authors explain their findings as follows:

  • when managers of companies decide whether goodwill has lost value, they bear in mind the costs that reporting an impairment loss may trigger.
  • managers ‘discount’ costs that will not occur until the future. In other words, managers perceive future costs as less costly than costs that must be paid immediately.
  • managers speaking strong-FTR languages perceive the future as more distant from the present than managers speaking weak-FTR languages do.
  • because of that perception, managers who speak a strong-FTR language discount more heavily the future costs that will result from reporting that goodwill has become impaired. So, those managers have more motive for delaying the recognition of impairment losses. That motivation makes them more likely to delay the reporting of a loss.

My reservations about the explanation

In the full version of this post on Accounting Miscellany, I explain my reservations about this explanation:

  • the explanation postulates that managers consider their own perception of how ‘psychologically distant’ future costs are when managers estimate whether goodwill has kept value or lost value. The paper does not explain how (or even whether) those perceptions can affect the impairment tests required by accounting standards (IFRS Standards).
  • although the paper reports that its findings are statistically significant, it does not discuss whether they are also economically significant. Knowing that a finding is statistically significant tells you that it probably wasn’t caused by chance alone. But knowing that a finding is economically significant tells you that it is large enough to deserve attention.    

Conclusions

I found this paper very stimulating and enlightening. The paper discusses whether (and how) one property of a language might affect one type of behaviour: testing whether goodwill has lost its value (has become ‘impaired’). That property is whether the language routinely marks verbs referring to the future.

The paper finds some thought-provoking evidence that speakers of a language marking verbs as future report impairment losses of goodwill more slowly than speakers of languages not marking verbs in this way. I have some reservations about the paper’s explanations for the findings.

I would love to see more research on how the properties of someone’s language affects that person’s behaviour. There is lots of scope for more research of this kind, in accounting and finance and in many other areas. I congratulate these authors for being among the pioneers of this kind of research. I look forward to seeing much more research by them and by many other researchers.

References

The Effect of Time Orientation in Languages on the Recognition of Goodwill Impairment Losses, by Ahmad Alshehabi, Hussein Halabi, Godfred Afrifa, in The International Journal of Accounting (2023), online ready at https://doi.org/10.1142/S1094406023500117

Chen, M. K. (2013). The effect of language on economic behavior: Evidence from savings rates, health behaviors, and retirement assets. American Economic Review, 103(2), 690–731
Online appendix at https://assets.aeaweb.org/asset-server/articles-attachments/aer/data/april2013/20111628_app.pdf

Dahl, Ö. (2000). The grammar of future time reference in European languages. In Tense and Aspect in the Languages of Europe (Vol. 6, pp. 309–328). Mouton de Gruyter Berlin.

Kim, J., Kim, Y., & Zhou, J. (2017). Languages and earnings management. Journal of Accounting and Economics, 63(2–3), 288–306. https://doi.org/10.1016/J.JACCECO.2017.04.001

Gotti, G., Roberts, S. G., Fasan, M., & Robertson, C. B. J. (2021). Language in economics and accounting research: The role of linguistic history. The International Journal of Accounting, 56(03), 2150015.

2 comments

Leave a comment

Your email address will not be published. Required fields are marked *