Self-Deception

"A foolish consistency is the hobgoblin of little minds, adored by little statesmen, and philosophers and divines" - Emmerson

The concept of self-deception has long been an attractive issue of study for philosophers interested in common language analysis. The attraction rises from the inherent paradox that arises from the state of being self-deceptive and the process of becoming self-deceived . Normally, one models what it means to be self-deceived after what it means to be deceived by someone else. It is often held that for you to deceive me, you must get me to believe wholeheartedly in a proposition that you know, or at least sincerely believe to be false. Following this model, then, self deception occurs when I can get myself to believe a proposition that I truly know is false. The paradox is that in order for self deception to be valid, an individual must hold at the same time, two contradictory propositions (p and not-p) and believe them both to be true. There have been various responses to this paradox from outright rejection of the concept, through to a rejection of the principle of contradiction. In this paper, I wish to consider 3 responses to the paradox of self-deception. 1) Partitioning strategies, 2) Redefinition Strategies, and 3) Para consistent Logic. I will claim that neither of these responses provides a sufficient explanation or solution to the problem. I will argue that, although there certainly seems to be valid cases of self-deception as we conceive it through common language, the analytic attempts have thus far been relatively unsuccessful.

Partitioning Strategies

Many philosophers have made sense of the self-deception paradox by partitioning the mind into separate logical subjects. In doing so, an individual can hold a proposition and the opposite of the proposition without being in a state of contradiction. If, for example, I hold the proposition p and not-p, and I believe them to be true, I am in a state of contradiction because I do not actually believe anything. However, if the mind consists of two or more parts, I can hold the proposition p (along with various reasons for p) within one part, ad the proposition not-p (along with reasons) in the other part. So long at these two parts are distinct, there will never be a contradiction.

One such practitioner is Ralphael Demos. He accepts the standard view that self-deception involves a contradiction:

"Self deception exists, I will say, when a person lies to himself, that is to say, persuades himself to believe what he knows is not so. In short, self-deception entails that B believes both P and not-P at the same time (1960, p.588).

But he divides the mind into distinct parts, separated by "levels of awareness":

"There are two levels of awareness possible; one is a simple awareness, the other awareness together with attending, nor noticing. It follows that I may be aware of something without, at the same time noticing it or focusing my attention on it" (p.593)

Therefore, for Demos, the deceiver is capable of simultaneously believing p and not-p "because he is distracted from the former" (p.594).

The problem with this view is stated succinctly by Herbert Fingarette (1969), who argues that Demo's own position is paradoxical. He points out that deception must be an intentional act. Therefore, if I am to deceive you, I must intend to pull the wool over your eyes. If I am merely wrong about a proposition I convince you to believe, but I sincerely believe that it is true, I have not deceived you - I have only demonstrated by ignorance. Similarly, if I am to deceive myself, I must do so intentionally. Therefore, the 2 propositions p (true) and not-p (false) and I notice not-p, but do not notice p (because not-p is distracting me) I am not self-deceived, merely ignorant of p. Furthermore, Finagarette argues, and individual cannot intentionally fail to notice p, because she has already noticed it: "it appears that it is just because he already appreciates the incompatibility of his beliefs that the self-deceiver 'deliberately ignores' the belief he abhors" (p.16). In this case, the individual would be arty to two contradictory beliefs, and the portioning strategy advanced by Demos has not helped us with the paradox.

Redefinition Strategies

Most common approaches to self-deception insist that the deception seen in self-deceptive situations is the same as the deception seen in interpersonal models. However, there are scholars who believe that the interpersonal models of deception are insufficient to describe self deception. As Mele (1987) states:

"One approach to resolving the paradox of self-deception is to abandon some pertinent features of interpersonal models of he phenomenon. Perhaps it is typically true that when A deceives B, A knows or believes the truth and intentionally gets B to believe a falsehood. But must the self-deceiver know or believe the truth and believe the negation of the true propositions?" (1987, p.8)

Mele answers in the negative, arguing that self-deception occurs because individuals prefer certain beliefs over others. These preferences direct individuals to manipulate the truth value of the respective claims. Thus,

". . . because, e.g., he (the subject takes a certain datum d to count against p, which proposition he wants to be the case, he may intentionally or unintentionally shift his attention away from d whenever he has thoughts of d; but to do this he need not believe that p is false"(1983, p.372).

Therefore, on this view, there is no requirement that the person believes not-p (regarded as true). I a mother wants her son to be good, but the relevant data would lead her to believe otherwise, she might manipulate the implications of the data, or the content of the data itself to represent that belief, but she does not also have to have a belief that her on is not-good the paradox of having two contradictory beliefs is thus avoided.

This is a nice distinction, but we are now left to wonder on the relationship between an individual perceiving a data set and the sae individual drawing conclusions based upon that data set. For Mele's distinction to hold, it seems that an individual would be able to perceive a data set, compare it with the desired proposition, and recognize that the two are inconsistent, ad then modify the dataset to match the desired belief, all without ever holding the contrary position. If a mother can recognize the data set about her son (i.e., that he steals from people and is continually violent), and she can recognize that these facts are inconsistent with her to have a good son, given Mele's logic, two cases are possible. If she manipulates this data "intentionally", she is aware of the not-good proposition and is in fact holding two contradictory beliefs. If, on the other hand, the data manipulation is done unintentionally ("presumably without awareness), it is being accomplished at a lower level of "awareness". If we claim that there are levels of awareness, we have another variant of a portioned mind theory, ad the problems we have already discussed crop up again. It seems that Mele's distinction is thus not valid.

Para consistent Logic Strategies

Up until now we have looked at theories that have had a goal of saving the concept or, self-deception from contradiction. This is because it is taken for granted by most that the formal logic principle of contradiction is as basic as you can get in philosophy; that is, all arguments should be free from contradiction. However, in recent logic theory, the foundational truth of the contradiction principle is being questioned by proponents of Para consistent logic . As it should be expected, this approach has been applied to the problem of self-deception. One of the primary examples of this approach can be found in da Costa and French (1990), who's explicit intention, is to:

"Liberate discussion of self-deception from the shackles of a purely classical logic, thereby permitting a separation of the more philosophical issues from those which might properly be described as "logical" (p.179).

Put simply, their approach will put forward "a Para consistent system which seems capable of accommodating contradictory beliefs" so that the paradox of self-deception no longer exists.

This is an interesting approach that is coupled with a full alternate logic system in the appendix of the paper. The details of logic systems are, for the purpose of this paper, less important than the rationale for adopting this approach. Da Costa and French begin by accepting the distinction between an unintentional inconsistency and an intentional inconsistency. They claim, quite reasonably, that a person may easily enter upon an unintentional inconsistency:

"We are not logically omniscient in the sense of being able to immediately deduce all the consequences of a given proposition that can be deduced by anyone that demands such omniscience is clearly asking too much"(p.185).

Unintentional inconsistencies are not of major interest, however. In fact, there presence in any belief system is a trivial claim; since they are unintentional they are not present in mind at all. What is interesting, for da Costa and French, is what happens when an individual becomes aware of an inconsistency in her belief system. In this case, the inconsistency becomes intentional, and:

"It might be argued that the only rational thing to do is to suitably rearrange one's set of beliefs with a view to eliminate the contradiction of the system"(p.186)

This is where the problem lies, for da Costa and French. They claim that because our standard belief systems are full of an incalculable number of inconsistent deductive chains, it may be impossible to eliminate an intentional inconsistency completely anyway. This is their reasoning for suggesting that the principle of contradiction be "weaken[ed] or abandoned altogether" (p.177)

They then turn to attack the principle of consistency on what they call pragmatic grounds. They ask us to consider an individual that has p and not-p in his mind. Both propositions, they claim reflect the process of reasoning over some sort of empirical data. There is usually lots of empirical data, and this data is usually able to be mobilized for different arguments. As such, an individual holding the propositions p and not-p might well have god reasons for both positions because the data supports them equally well. Thus,

"The removal of inconsistency in this manner may therefore be a practically impossible, or near impossible, undertaking. One might have good reason to hold both of a pair of contradictory beliefs. Thus, for example, the same empirical evidence might equally support the same two conflicting theories, or different, but equally acceptable, pieces of evidence might support two contradictory propositions within a given theory . . . "(p.186).

In this view, a mother who sees her son might view the facts (i.e., "that he is kind towards a senior") and hold the position that her son is good. On the other hand, she might view another set of facts (i.e., that he steals and is violent) and hold the proposition that her son is not-good. Da Costa and French believe that this consists in holding two contradictory propositions; and, more importantly, they hold that this inconsistency is inherent for human beings trying to understand the complex world. As this is par for human existence, we should not be rigid on the logic of contradiction and perhaps a logic that abandons this principle is in line with how we process information anyway (p.190). With this new logic that accepts contradictions, self-deception would be a valid concept, despite the fact that it depends on contradiction.

This argument sounds acceptable at first glance. However, if considered, I think there is something wrong with it. The main thrust of the argument is that human beings naturally hold inconsistent thoughts and there is no ridding ourselves of inconsistency. As such, we are setting our goals too high to ask for consistent thoughts by utilizing the principle of contradiction. This argument depends; it seems, on two things. First, it depends on a set of facts that are so infinitely complex that we can never determine them all (or know when we have); and second, it depends on individuals "collecting" distinct propositions based upon differing data.

Although the first dependency may not be self-obvious, it is sound enough to leave it be here. However, it is not at al clear that individuals may draw distinct propositions based on differing data. That every month will draw two inconsistent propositions from the fact that "my son steals" and "my son is kind to seniors" require much more argument than is present. Rather than drawing the inconsistent -distinct- propositions "my son is god" and "my son is not good", it seems equally likely that the conclusion might be the consistent claim that "my son is imperfect". Or, assuming that an individual might value private property more than a principle of kindness, these two facts might still manifest themselves into the propositions "my son is not-god". I suggest here that da Costa and French are relying upon a theory of data collection and synthesis that they have certainly not argued for. It is not clear that humans are naturally inconsistent at al, except in the sense that they've defined it.

Conclusions: Is Self-Deception a Viable Concept?

When I first began this paper, I believed self-deception to be a viable notion, resting not on a contradiction between two contradictory propositions, but on a distinction between the real state of affairs and an individual's Ideal state of affairs. I am now uncertain that this claim can be made. After all, would this not still be a contradiction in beliefs? For instance, a contradiction in the belief f what is real and what one wish to be real? Perhaps there is some room to move with this claim because a wish might not hold the same epistemic status that a belief would (i.e., the status of a proposition). Unfortunately, I am unable to pursue this here. I am certain that the issue of self-deception is more complex than I initially had thought. I do not find the solutions to the problem presented here particularly satisfying. However, I hope I have shown with some clarity why they are so. Perhaps there are further views that I have not yet considered, and I should like to pursue this issue further at some point in the future.

Bibilography

Arruda, A.I, (1980.n "A Survey of Paraconsistent Logic", n A.I Arruda, R. Chuaqui, and N.C.A da Costa, eds. Mathematical Logic in Latin America (North Holland, 1980). P.1-41
daCosta, N.C.A & French, S (1990). "Belief, Contradiction and the Logic of Self-Deception", American Philosophical Quarterly, Vol. 27, p.179-197.
Demos, R. (1960), "Lying to Oneself". Joural of Philosophy Volume 57, p.588-595
Finagarette, H. (1969) Self-Deception. London: Routeledge and Kegan Paul.
Mele, A. (1983), "Self-Deception". Philosophical Quarterly vol. 33, p.365-377.
Mele, A (1987), "Recent Work on Self Deception", American Philosophical Quarterly, vol. 24, Jan 1987.
Priset, G., Routely, R & Norman, J. (eds) (1989). Paraconsistgent Logic: Essays on the Inconsistent (Munich: Philosophia Verlag; 1989).

For a discussion of the state / process paradoxes of self-deception see Mele (1987)
There are many other partitioning arguments that differ from Demo's, but the length of this paper precludes their mention.
There are a number of examples of Para consistent logics. See Arruda, A. (1980) and Priest (1989) for examples. They all seem to follow what might be called a "naturalist" approach.
And, to be quite honest, I would be deceiving myself if I thought I understood the system that is presented.
This is probably a tendentious claim. If an intentional consistency is removed from the system, it may still conflict with other claims. But if these claims are unknown, they would be unintentional consistencies. If they are unintentional consistencies, we would not be aware of them anyway, so how would our intentional system be affected?

Canadian Electoral Reform

The 1993 Canadian federal election has renewed many past cries for electoral reform. The complete devastation of the previously strong Progressive Conservative party, as well as the predominance of the strictly regionalized parties (i.e., Bloc Quebecois and Reform); have left many repulsed by the inadequacies of the first-past-the-post (FPTP) system. Despite the quite obvious dissatisfaction, however, electoral reform has not been implemented, nor does such reform seem promising for the future. It seems, therefore, that there are those who think the present system best, and thus change unnecessary. This paper, will analyze the current electoral system, illuminating its many faults. It will advocate the implementation of Proportional Representation (PR) - the other major - system as well as deal with various criticisms levelled against it by FPTP supporters.

Under a FPTP electoral system, the country is divided into numerous regions, or constituencies. From each constituency, one candidate is elected to represent its electorate in government. This method is labelled a "single member system" - only on representative from each constituency is elected. Candidates "win" an election by achieving a "plurality" f the popular vote. This means, simply, that the candidate with the most votes wins. It is quite important to note that a plurality of votes need not be a majority. For example, if 20% of the popular vote was the highest percentage in a constituency, the party will win the election ad proceed to govern even though a full 80% of the population did t support that candidate. This illuminates an interesting quirk in the FPTP system. Theoretically, if this statistic was amplified nation wide, the party would govern with less than 25% of the population supporting their campaign. I a country where democracy is assumed to equal majority rule, this situation seem antithetical.

Proportionality of Votes to Seats

These are the basic institutional characteristics of the Canadian FPTP system. In order to acquire a complete understanding, however, the practical implications of these characteristics must now be explained.

By allowing the party "first-past-the-post" to win, there is often great injustice done to other parties. This can be seen quite clearly in the 1993 Prince Edward Island provincial election. The Liberal party, headed by Catherine Callebeck, won 97% of the seats with only 55% of the vote. This left the Conservative party, representing a full 40% of the population with only one seat, and the NDP, representing 5%, with not even that many (Hiemstra, In Charlton & Barker (194:281). Quite simply, this is because in each of the 32 constituencies (with the exception f the one Tory constituency) the Liberals had more votes - o mater how small the difference was.

Concurrently, in the 1993 federal election, the National party, running cross-country, received a significant percentage of the vote, yet no representation whatever. It is quite possible to have a party victory with a 0.1% advantage. It might give them complete governing power, while the loser receives a clap on the back and an "A+" for effort.

This shows quite clearly that many parties do not receive the power given to them by the people. Relatively small, non-traditional, nationally based parties are almost always negatively affected by this situation (Cairns 1968:64). These parties, as they often offer les traditional policy platforms, ad do not have large resources or the reputations, rarely receive enough electoral support to attain the most votes in any constituency. This means that while they hold a significant percentage of the popular vote, they may receive no governing power. In 935, for example, the CCF received 9% of the vote in the Alberta provincial election, yet received no seats in the legislature. In a country where the performance of democracy is held in high esteem, it is interesting that the electoral system serves to mute the voice of those who refuse to propagate the hegemon, or status quo.

Regional Disparity

The other major implication of FPTP is a take-off from the first. Realizing the difficulty in obtaining enough national support to govern, most parties will "wisely" choose to run candidates only where there is a chance for electoral victory (Cairns: 1968:62). After all, if the purpose of running for election is to be elected, and the electoral system discriminates against (small) national arties (Cairns 1968:59-62), the obvious strategy is to run were regional support is high - thus ensuring at least some power rather than none for the same effort.

This strategy can clearly be seen within the Tory party. Historically, Quebec has been the Tory electoral nemesis. They have failed miserably to gain significant Quebec support for their right-wing policies. In 1953, chief strategist Gordon Churchill advised his party to "reinforce success, not failure" (Russell, 1993); that is, to concentrate n the West and forget Quebec.

What occurs, therefore, is regional based parties sprouting up rapidly. The 1993 Reform party as well as the Bloc Quebecois provide excellent examples of such parties. The Bloc ran only in Quebec, yet found enough support (16.05%) to launch it to official opposition status (54 seats). Reform, which had all but one of its MPs elected from Western provinces (13.72%), now holds 52 seats.

Although on may not see the disadvantage of this practical implication, it becomes obvious when we examine actual legislation. The controversial National Energy Policy serves as a prime example. In 1980, the Liberal party failed to elect a member West of Winnipeg. Utilizing their majority, however, they implemented this legislation, ignoring completely the outrage stemming from the West. The Eastern provinces, on the other hand, were emphatically supportive f it. The result was quite clearly an increase in regional animosity; the West resenting the East for receiving preferential treatment.

In an age where national unity is living a precarious existence, it is hardly advantageous to have these animosities. In fact, it may, in retrospect, be logical to accuse FPTP of developing this animosity. Propagating this system might eventually lead Canada down the dark and dismal path to failure - that is, regional separation.

It may seem that a system under which such blatant injustices occur would have few advocates. To assume such simplicity would be rather naive, however. While most scholars to a greater or lesser extent agree with these criticisms, they argue that FPTP is the best available alternative. Proportional Representation (PR) - the other major system - has problems that make the trade well worth it.

The most commonly advocated PR system is the party list system (see, for example, Hiemstra, ibid). Under this system, voters would not vote for a specific candidate, but a party. Parties would prepare lists where members are "franked" n order of electoral importance. There are single constituencies, rather than provinces being treated as a constituency. Most importantly, the percentage of votes a party receives is translated directly into the percentage of seats it receives. For example, if a province has 100 seats and it receives 50% of the vote, it would get t50% of the seats. These seats would then be filled by the first 50 members on the list.

This system would eliminate the FPTP representational problems. No longer would traditional, regionalized parties be favoured. A party would receive exactly the same percentage of governing power as popular vote. If a party would receive 5% of the vote, it would receive 5 of the seats. Furthermore, a national unity would be promoted. After all, under this system a party would not accrue any benefit in regionalization. It would be illogical for ay party to not run nationally.

Although PR seems to eradicate the evils of FPTP in ne fell swoop, FPTP supporters remain unsatisfied. The most fundamental flaw highlighted is its propensity to generate minority governments. One f the major so-called advantages of the FPTP system is that its tendency to generate stable, majority governments is higher. The reasoning being that a stable government is able to provide greater legitimacy and better government.

Admittedly, PR does produce majority governments. After all, a party under any system very rarely receives the 50% + 1 vote constituting a majority. This would probably occur less often under PR as more parties would be encouraged to run. That this is a disadvantage, however, is uncertain.

Minority coalitions may indeed be more conducive to democracy than its FPTP counterpart. After all, if the governing parties must work collectively on legislation, it seems that a wider range of public interest will be served with the more popular vote that is represented. Rather than a government (who may represent little of the population) forcing through legislation against the will of many citizens (i.e., Brian Mulroney's G.S.T tax), there must be rational compromises made. This in itself seems quite democratic.

In addition, we may examine a recent comparative study conducted by Lijphart (1994: 1-17). He analyzes the conventional statement that PR systems give better representation that PR systems give better representation at the cost of poor government stability and performance. His study clearly validates the former claim. He demonstrated that PR counties are more receptive to minority opinion, more egalitarian, generate more effective policy and have a greater percentage of voter turnout - all measure of representation. The latter claim, however, was found to be false. No statistically significant difference in economic growth, inflation, unemployment, or violent political strife - all measures f government performance - was found. If anything, he claims PR governments perform this function better too! Therefore, PR seems to generate greater democracy without giving up government performance.

Many scholars will, however, still retain their preference for stability - that is, a long lasting government. A minority, they claim, will fall quickly due to irreconcilable party differences. This will result in frequent elections, and thus constant political turmoil.

It is true that, in Canada, minority governments have tended to fall quickly. As Hiemstra (1994:283) argues, however, this maybe the result f the FPTP incentive rather than any inherent failure in minority governments. Under FPTP, parties realize that a small shift in popular vote may give them the majority they desire, so they tend not to negotiate well and the minority collapses. Under PR, they will have no such incentive, thus cooperation would be more likely.

This argument fails to explain countries like Italy, admittedly. Italy, however, has far different political history than Canada. Canada has been a stable democracy since it was founded. There is no logical reason to assume that Canada would not resemble the stability of the Netherlands, for example.

Finally, although FPTP systems are presumed to generate majority governments, history proves this assumption fallacious. Since 1921, there have been a total of 21 elections, 8 of which have been minority governments. This is a surprisingly "poor" record (Reports f the Chief Electoral Officer). Even though the average duration f these governments were 1.73 years, it is fair to say that Canadian democratic stability remains unquestioned.

As we may now see, the fact that PR tends to produce minority governments is hardly a disadvantage. If anything, t might be a decisive advantage. Not only do minority systems provide netter government (as Lijphart demonstrates), but they do so more democratically.

Conclusion

Although this paper has been pro-PR, one should not be misled. PR systems do have faults. The party list system described above, while commendable, is subject to valid criticism. After all, who makes the list and who decides what order candidates are placed on it? The party leader, of course. Obviously higher numbered candidates have a greater chance of becoming elected than lower candidates. This might practically result in patronage politics - hardly the democratic spirit.

Critics cite this fact routinely to counter the benefits of PR (Barker, 1984:295). What must be realized, however, is that much of the same patronage occurs under the present system anyway. In ay election, the party leader attracts the most media attention. All candidates would therefore benefit greatly from having their leader appear publically in their campaign. Not only would this bring greater media publicity, but also an added respect. Public appearances are at the complete discretion f the leader. Obviously, the leader will rather support a "loyal" candidate than a "disloyal" one.

Moreover, it is the party leader who is in control f the campaign funds. IF candidates are disloyal, perhaps they will not receive adequate funding, thus reducing chances of success.

To say, therefore, that FPTP systems eliminate this problem is ridiculous. The only difference is under FPTP, patronage is covert, while under PR it is relatively overt. The difference is only a matter f kind. It seems, perhaps unfortunately, that patronage is a practice is a practice irremovable from the political process. In this case, blame the person, not the policy. We must make a priority decision. Must the possibility of PR reform be eliminated because it does not eradicate one flaw?

We have finally reached the end of our journey. While this paper has analyzed many characteristics of FPTP and PR, it should not be considered conclusive. There are other minor criticism advanced against PR, but we will not deal with them here. This analysis, however, focuses on the most crucial aspects of electoral reform. It demonstrates logically and empirically the plausibility, despite the criticisms, of the implementation of PR into Canadian politics.

Not only would it eradicate the present representational problems, but it may provide us with better government. After all, as most great innovations are the result of cooperative dialogue, innovative policy-making may not be far to follow.

Bibilography

Barker, Paul, "Voting for Trouble", in Charlton, Mark & Barker, Paul (Eds.). Crosscurrents: contemporary Political Issues 2nd Ed. Nelson Canada. Pg 292-301.
Cairns, Alan (1968). "The Electoral System and the Party System in Canada", in Canadian Journal of Political Science, 1968, pg. 55-80.
Cassidy, Michael (1991). Ed. Democratic Rights and Electoral Reform in Canada Volume 10, Dundurn Press, Toronto & Oxford.
Hiemstra, John, I Charlton, Mark, ad Barker, Paul (eds.) "Getting What you Vote for". Crosscurrents: Contemporary Political Issues, 2nd Ed. Nelson Canada, pg 280-291.
Irvine, William, P. (1979). Canada Need A New Electoral System? Institute of Intergovernmental Relations.
Lijphart, Arend (1994), "Democracies: Forms, Performances & Constitutional Engineering" European Journal of Political Research, Jan, 1994.
Reports of the Chief Electoral Officer
Russell, Frances, "There are Better ways to Vote, Winnipeg Free press, Nov 5, 1994