Computational MetaEthics
 1.6k Downloads
 2 Citations
Abstract
It has been argued that ethically correct robots should be able to reason about right and wrong. In order to do so, they must have a set of do’s and don’ts at their disposal. However, such a list may be inconsistent, incomplete or otherwise unsatisfactory, depending on the reasoning principles that one employs. For this reason, it might be desirable if robots were to some extent able to reason about their own reasoning—in other words, if they had some metaethical capacities. In this paper, we sketch how one might go about designing robots that have such capacities. We show that the field of computational metaethics can profit from the same tools as have been used in computational metaphysics.
Keywords
Automated moral reasoning Computational metaethicsIntroduction
A continually growing number of computers/robots is being deployed on the battlefield, in hospitals, in law enforcement, in electronic business negotiation and other ethically sensitive areas. It is desirable that the computers/robots in these areas behave in an ethically correct manner. It might be argued that they can only be guaranteed to do so if they can reason about right and wrong on the basis of a set of moral standards: only “explicit ethical agents” (as they have been called; see Moor 2006) can be expected to behave in an ethically correct manner. However, sets of moral standards can be inconsistent, incomplete, or inappropriate in view of other sets of standards; it would therefore desirable that robots equipped with such standards were to some extent able to reason about them—in other words, if they had some capacity for metaethical reasoning.
These considerations suggest that when one wants to design ethically correct robots one should not only explore the field of automated moral reasoning, but also the field of automated moral metareasoning, in the sense of reasoning about moral reasoning. The development track that we foresee is roughly as follows: informal moral reasoning (reasoning about obligations, permissions and prohibitions)—formal moral reasoning (deontic logic)—automated moral reasoning (computational deontic logic)—automated reasoning about automated moral reasoning (computational metaethics).
Along this development track, the area of automated moral reasoning has already received some attention. Computational deontic logic has been discussed in the context of computersupported computer ethics (Hoven and Lokhorst 2002). This work has been used in books on the design of moral machines that can distinguish right from wrong (Wallach and Allen 2008) and engineering autonomous (unsupervised) moral robots that are capable of carrying out lethal behavior (Arkin 2009). Similar research has been described in papers on moral reasoning by ethically correct robots (Arkoudas et al. 2005; Bringsjord et al. 2006).
However, the subject of automated moral metareasoning has not received much attention so far. We intend to set some cautious first steps in this area. We are interested in machines that reason about moral reasoning—metaethical robots—for two reasons. First, it is an intellectual challenge to think about how one might go about constructing such machines, which might be called philosophical in the sense that philosophers (inspired by Carnap 1937) are sometimes fond of saying: “anything you can do, I can do meta.” Second, it seems important for practical applications. Ethical robots need ethical standards. However, these standards may need examining, either during the design process or in the deployment stage. To the extent that metaethical reasoning can be delegated to computers/robots, the results will be both more easily obtainable and more reliable than they would be if this reasoning were left to the humans who design or deploy them.
We will proceed on the assumption that one can use some of the same computational tools in computational metaethics as have been used in computational metaphysics (Fitelson and Zalta 2007) and in metareasoning about different systems of modal logic (Rabe et al. 2009). We make this assumption because we have little reason to believe that metaethics is essentially different from (let alone more difficult than) metaphysics or metamathematics.
Design of a MetaEthical Robot
Nine Modules
 1.
Seven nondeontic logical modules.
 2.
One deontic module.
 3.
One metalogical module.
 1.Module \({\varvec{\mathfrak{S}}_{\rm T}^{\circ}}\) implements system \({\varvec{\mathfrak{S}}_{\rm T}^{\circ}}\), which we define as the weakest traditionally strict classical logic closed under strict modus ponens. \({\varvec{\mathfrak{S}}_{\rm T}^{\circ}}\) has the same language as the propositional calculus, except that there is an additional unary connective square, read as “necessarily,” and the definitions \(A\rightarrow B\mathop{=}\limits^{\hbox{df}}\square(A\supset B)\) and \(A\leftrightarrow B\mathop{=}\limits^{\hbox{df}} (A\rightarrow B)\,\&(B\rightarrow A)\), where → is strict implication, \(\supset\) classical material implication, and \(\leftrightarrow\) strict equivalence. The axioms and rules of inference are as follows (Chellas and Segerberg 1996):

PL The set of all tautologies.

\(\square{\text {PL}}\;\{{\square{A}:A \in {\text {PL}}}\} \).

US Uniform substitution.

\(\hbox{RRSE}_{\rm T}\) From \(A\leftrightarrow B\) and F to infer \(F^{A/B}\), where \(F^{A/B}\) is the result of replacing one occurrence of A in F by B.

MP From A and A⊃ B to infer B.

SMP From A and \(A\rightarrow B\) to infer B.
All Lewis systems of strict implication (as studied in Zeman 1973, for example) are traditionally strict classical logics closed under strict modus ponens. Some of these systems and the relations between them are as follows:$$ {{\mathfrak{S}}}_{\rm T}^{\circ} \subset {\bf S0.9}^\circ \subset \left\{ \begin{array}{l} {\bf S0.9}\\ {\bf S1}^{\circ}\\ \end{array} \right\} \subset {\bf S1} \subset {\bf S2}\subset {\bf S3}\subset {\bf S4}\subset {\bf S5}. $$ 
 2.Module \({\bf \L}_{\aleph_0}\) implements Łukasiewicz’s infinitevalued logic \({\bf \L}_{\aleph_0}\), which has the following axioms and rules (Malinowski 2001):
 1
\(A\rightarrow (B\rightarrow A)\)
 2
\((A\rightarrow B)\rightarrow ((B\rightarrow C)\rightarrow (A\rightarrow C))\)
 3
\(((A\rightarrow B)\rightarrow B)\rightarrow ((B\rightarrow A)\rightarrow A)\)
 4
\((\sim A \rightarrow \sim B)\rightarrow (B\rightarrow A)\)
 5
\(((A\rightarrow B)\rightarrow (B\rightarrow A))\rightarrow (B\rightarrow A)\)
 6
From A and \(A\rightarrow B\) to infer B.
Definitions: \(A\lor B\mathop{=}\limits^{\hbox{df}} (A\rightarrow B)\rightarrow B\), \(A\,\& B\mathop{=}\limits^{\hbox{df}}\sim(\sim A\lor \sim B)\), \(A\leftrightarrow B\mathop{=}\limits^{\hbox{df}}(A\rightarrow B)\,\&(B\rightarrow A)\). \({\bf \L}_{\aleph_0}\) is the weakest of all Łukasiewicz systems, in the sense that A is a theorem of \({\bf \L}_{\aleph_0}\) if and only if A is a theorem of all finitevalued Łukasiewicz calculi \({{\bf \L}_n, n\geq 2, n\in{\mathbb{N}}}\). Formally, if Th(S) is the set of theorems of S, then \({Th({\bf \L}_{\aleph_0}) = \bigcap\{Th({\bf \L}_n)\colon n\geq 2, n\in{\mathbb{N}}\}}\) (Malinowski 2001). These systems are nowadays popular in the field of fuzzy logic.
 1
 3.Module H implements Heyting’s system of intuitionistic logic H, which has the following axioms and rules (Dalen 2001):Intuitionistic logic is related to minimal logic, which we will not study separately.
 1
\(A\rightarrow (B\rightarrow A)\)
 2
\((A\rightarrow (B \rightarrow C)) \rightarrow ((A\rightarrow B)\rightarrow (A\rightarrow C))\)
 3
\((A \,\& B) \rightarrow A\)
 4
\((A \,\& B) \rightarrow B\)
 5
\((A\rightarrow (B \rightarrow (A \,\& B))\)
 6
\(A\rightarrow (A\lor B)\)
 7
\(B\rightarrow (A\lor B)\)
 8
\((A\rightarrow C)\rightarrow ((B\rightarrow C)\rightarrow ((A\lor B)\rightarrow C))\)
 9
\(A \rightarrow (\sim A \rightarrow A)\)
 10
\((A \rightarrow B) \rightarrow ( (A\rightarrow \sim B) \rightarrow \sim A)\)
 11
From A and \(A\rightarrow B\) to infer B.
 1
 4.Module R implements relevant system R, which has the following axioms and rules (Dunn and Restall 2002):Relevance logic is a predecessor of linear logic, which we will not study separately.
 1
\(A\rightarrow A\)
 2
\((A\rightarrow B)\rightarrow ((C \rightarrow A)\rightarrow (C\rightarrow B))\)
 3
\((A\rightarrow (B \rightarrow C)) \rightarrow ((A\rightarrow B)\rightarrow (A\rightarrow C))\)
 4
\(A\rightarrow ((A\rightarrow B)\rightarrow B)\)
 5
\((A \,\& B) \rightarrow A\)
 6
\((A \,\& B) \rightarrow B\)
 7
\(((A\rightarrow B)\,\& (A\rightarrow C)) \rightarrow (A\rightarrow (B \,\& C))\)
 8
\(A\rightarrow (A\lor B)\)
 9
\(B\rightarrow (A\lor B)\)
 10
\(((A\rightarrow C)\,\& (B\rightarrow C))\rightarrow ((A\lor B)\rightarrow C)\)
 11
\((A\,\&(B\lor C))\rightarrow ((A\,\& B)\lor (A\,\& C))\)
 12
\((A\rightarrow\sim B)\rightarrow (B\rightarrow \sim A)\)
 13
\(\sim\sim A\rightarrow A\)
 14
From A and \(A\rightarrow B\) to infer B.
 15
From A and B to infer \(A\,\& B\).
 1
 5.
Module RM implements relevant system RM, which is defined as R plus axiom \(A\rightarrow(A\rightarrow A)\).
 6.
Module RM3 implements relevant system RM3, which is defined as R plus axioms \((\sim A\,\& B)\rightarrow (A\rightarrow B)\) and \(A\lor (A\rightarrow B)\) (Dunn and Restall 2002).
 7.
Module KR implements relevant system KR, which is defined as R plus axiom \((A\,\&\sim A) \rightarrow B\) (Dunn and Restall 2002). Systems R, RM, RM3 and KR are related as follows: \({\bf R} \subset {\bf RM} \subset {\bf RM3}\), \({\bf R} \subset {\bf KR}\), \({\bf RM} \nsubseteq {\bf KR}, {\bf RM3} \nsubseteq {\bf KR}, {\bf KR} \nsubseteq {\bf RM}, {\bf KR} \nsubseteq {\bf RM3}\). System KR is famous for being the first relevance logic that was shown to be undecidable.

OA: it is obligatory that A.

u: that which is unconditionally obligatory.

\({\circledR}_{{A}}{{B}} \mathop{=}\limits^{\hbox{df}} A \rightarrow O B: A\) requires B.
 1
\(({\circledR}_{{A}}{{B}}\,\& (B\rightarrow C) )\rightarrow {\circledR}_{{A}}{{C}}\)
 2
\(({\circledR}_{{A}}{{B}}\,\& {\circledR}_{{A}}{{C}})\rightarrow {\circledR}_{{A}}{{(B\,\&\, C)}}\)
 3
\({\circledR}_{{A}}{{B}}\leftrightarrow O (A\rightarrow B)\)
 4
O u
 5
\(\sim{\circledR}_{u}\sim u\)
 1
Formula T1 \(A\rightarrow O A\) (if A is the case, then A is obligatory) is undesirable.
 2
Formula T2 \(O A\rightarrow A\) (if A is obligatory, then A is the case) is undesirable.
 3
Formula T3 \(O(A \lor B)\rightarrow (O A\lor O B)\) (if it is obligatory that either A or B, then either A is obligatory or B is obligatory) is undesirable.
 4
Formula D \(O A\rightarrow \sim O\sim A\) (if A is obligatory, then the negation of A is not obligatory) is desirable.
The nine modules we have mentioned embody insights that were available around 1930. We confine our attention to these ideas because logicians have put forward so many proposals since then that we cannot possibly take all of them into account.
Flexibility of Architecture

The nondeontic modules can be turned on and off.

The deontic and metaethical modules are fixed.
Modus Operandi
In order to evaluate the logical modules, module Meta is equipped with theorem provers and model generators (which produce counterexamples to formulas). These programs work either concurrently on one processor or in parallel on multiple processors. If the logic under examination is decidable, then given a formula A, either a theorem prover or a model generator will produce a result (if the programs work as advertised).
The scores for the four reference formulas T1, T2, T3 and D determine whether module Meta accepts or rejects the logic under examination. Module Meta accepts the logic as soon as each of the undesirable formulas (T1, T2 and T3) is shown to be invalid by some countermodel generator and the desirable formula D is shown to be derivable by some theorem prover. Module Meta rejects the logic it as soon as some undesirable formula (T1, T2 or T3) is shown to be derivable by some theorem prover or the desirable formula D is shown to be invalid by some countermodel generator. Module Meta remains in a state of indecision as long as the logic under examination has not been accepted or rejected. This state of indecision may last forever in the case of R and KR because these systems are undecidable. This would be fatal for a robot on the battlefield or in a hospital, but we have not encountered this situation in the cases we are examining.
MetaEthical Reasoning Process
In the example that we are going to describe, module Meta contains one theorem prover, namely Prover9 (McCune 2008), and two countermodel generators, namely Mace4 (McCune 2008) and MaGIC (Slaney 2008). The latter program is especially suitable in the case of relevant systems, but it was useful for the refutation of T3 in H as well.
Formula  D1–D5 plus  

\({\varvec{\mathfrak{S}}_{\hbox {T}}^{\circ}}\)  \({\bf \L}_{\aleph_0}\)  H  R  RM  RM3  KR  
\({\bf T1}\quad A\rightarrow O A\)  +  +  +  −  −  −  − 
\({\bf T2}\quad O A\rightarrow A\)  +  +  −  −  −  +  − 
\({\bf T3} \quad O(A\lor B)\rightarrow(O A\lor O B)\)  +  +  −  −  +  +  − 
\({\bf D} \quad O A\rightarrow\sim O\sim A\)  +  +  +  −  −  +  + 
The table shows that only KR is acceptable in view of the metaethical standards employed by the robot.
Conclusion
We have shown that computational metaethics is to some extent feasible, using current, offtheshelf, generic, freely available opensource software.

Different logical systems, both deontic and nondeontic.

Different theoremprovers and modelgenerators, for example the awardwinning Vampire and Paradox (Rabe et al. 2009).

Different metaethical criteria, for example, acceptability and unacceptability of alternative deontic principles. Freedom from “isought fallacies” would be an example. Isought fallacies are formulas of the form \(A\rightarrow O B\) or \(A\rightarrow\sim O B\), where A and B contain no occurrences of O and u. It can be proven that R plus Mally’s axioms is the only system on our list that avoids such fallacies, but we doubt whether the computer is clever enough to see this.

Different ways of reasoning about (moral) reasoning, for example in terms of computational complexity and computational tractability.

More expressive languages, for example languages with perception, knowledge, action, multiple agents, strategies, intention and time (see Hoven and Lokhorst 2002 and Lokhorst and Hoven 2011 for further discussion). In this context, it is interesting to know that large parts of multimodal correspondence theory have recently been mechanized, which makes it easier to study the interactions of modalities such as belief, seeing to it that, possibility, obligation and temporal notions (Georgiev et al. 2006).

On the fringe of logic and beyond logic: probabilistic reasoning, moral belief revision, the moral frame problem, nonmonotonic reasoning and noninferential moral judgment, for example casebased judgment and pattern recognition with neural networks (see Gärdenfors 2005 for more about these topics).
In other words, the design of a fully fledged metaethical robot is still a long way off. However, developments in the recent past suggest at least one direction in which we can set out on the long road that lies ahead.
Notes
Open Access
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
References
 Anderson, A. R. (1967). Some nasty problems in the formal logic of ethics. Noûs, 1, 345–360.CrossRefGoogle Scholar
 Arkin, R. C. (2009). Governing lethal behavior in autonomous robots. Boca Raton, Fl.: Chapman & Hall.CrossRefGoogle Scholar
 Arkoudas, K., Bringsjord, S., & Bello, P. (2005). Toward ethical robots via mechanized deontic logic. In Machine ethics: Papers from the AAAI fall symposium, Technical Report FS–05–06. Menlo Park, Cal.: AAAI Press.Google Scholar
 Bringsjord, S., Arkoudas, K., & Bello, P. (2006). Toward a general logicist methodology for engineering ethically correct robots. IEEE Intelligent Systems, 21(4), 38–44.CrossRefGoogle Scholar
 Carnap, R. (1937). The logical syntax of language. London: Routledge and Kegan Paul.Google Scholar
 Castañeda, H.N. (1981). The paradoxes of deontic logic: The simplest solution to all of them in one fell swoop. In R. Hilpinen (Ed.) New studies in deontic logic. Dordrecht: Reidel.Google Scholar
 Chellas, B. F., & Segerberg, K. (1996). Modal logics in the vicinity of S1. Notre Dame Journal of Formal Logic, 37, 1–24.MathSciNetzbMATHCrossRefGoogle Scholar
 Dunn, J. M., & Restall, G. (2002). Relevance logic. In D. Gabbay & F. Guenthner (Eds.), Handbook of philosophical logic (2nd Edn., Vol. 6). Dordrecht: Kluwer.Google Scholar
 Fitelson, B., & Zalta, E. N. (2007). Steps toward a computational metaphysics. Journal of Philosophical Logic, 36, 227–247.MathSciNetzbMATHCrossRefGoogle Scholar
 Gärdenfors, P. (2005). The dynamics of thought. Dordrecht: Springer.Google Scholar
 Georgiev, D., Tinchev, T., & Vakarelov, D. (2006). Sqema (an algorithm for computing firstorder correspondences in modal logic) version 0.9.8, September 2006. http://www.fmi.unisofia.bg/fmi/logic/sqema/.
 Lokhorst, G. J. C. (2008). Mally’s deontic logic. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy. plato.stanford.edu/entries/mallydeontic/.Google Scholar
 Lokhorst, G. J. C. (2010). Where did Mally go wrong? Lecture Notes in Artificial Intelligence, 6181, 247–258.Google Scholar
 Lokhorst, G. J. C. (2011). Where did Mally go wrong? Journal of Applied Logic (accepted).Google Scholar
 Lokhorst, G. J. C., & van den Hoven, M. J. (2011). Responsibility for robots. In P. Lin, K. Abney & G. Bekey (Eds.) Robot ethics: The ethical and social implications of robotics. Cambridge, Mass.: MIT Press. (Forthcoming in 2011).Google Scholar
 Malinowski, G. (2001). Manyvalued logics. In L. Goble (Ed.), The Blackwell guide to philosophical logic. Oxford: Blackwell.Google Scholar
 McCune, W. (2008). Prover9 and Mace4 version 200811A, November 2008. http://www.cs.unm.edu/~mccune/prover9/.
 Moor, J. H. (2006). The nature, importance, and difficulty of machine ethics. IEEE Intelligent Systems, 21(4), 18–21.CrossRefGoogle Scholar
 Rabe, F., Pudlák, P., Sutcliffe, G., & Shen, W. (2009). Solving the $100 modal logic challenge. Journal of Applied Logic, 7, 113–130.MathSciNetzbMATHCrossRefGoogle Scholar
 Slaney, J. K. (2008). MaGIC (Matrix Generator for Implication Connectives) version 2.2.1, November 2008. users.rsise.anu.edu.au/~jks/magic.html.Google Scholar
 Van Dalen, D. (2001). Intuitionistic logic. In L. Goble (Ed.), The Blackwell guide to philosophical logic. Oxford: Blackwell.Google Scholar
 Van den Hoven, M. J., & Lokhorst, G. J. C. (2002). Deontic logic and computersupported computer ethics. Metaphilosophy, 33, 376–386. Reprinted in J. H. Moor & T. W. Bynum (Eds.) (2002). CyberPhilosophy. Oxford: Blackwell.Google Scholar
 Wallach, W., & Allen, C. (2008). Moral machines: Teaching robots right from wrong. Oxford: Oxford University Press.Google Scholar
 Zeman, J. J. (1973). Modal logic: The LewisModal systems. Oxford: The Clarendon Press. http://www.clas.ufl.edu/users/jzeman/modallogic/.