The Concise Oxford Dictionary of Politics (225 page)

(1632–94)
Jurist whose main contribution was to international law. Born near Chemniz, Saxony, he studied law at Leipzig and Jena, and taught at Heidelberg and Lund. He was imprisoned by the Danes because of his contact with the Swedish ambassador, whose sons he tutored in Copenhagen. While in prison he wrote
The Elements of Universal Jurisprudence
(1660). His main work, written at Lund, was
On Natural Law and the Law of Nations
(1670). He also wrote
On the Duty of Man and of the Citizen
(1671), and
On the Relation between Church and State
(1686).
Pufendorf followed
Grotius
for the most part, but interpreted
ius gentium
more positivistically, thus breaking from the Aristotelian tradition. He introduced elements of
Hobbes's
conventional, contractual idea, without carrying self-interest as far as Hobbes did. For him, as for Grotius, a firmer, more rational basis for a political society was necessary. In keeping with his positivistic approach, he came close to
Rousseau's
notion of the
general will
or the state as a moral individual whose will is the resultant when individual citizens' wills have cancelled each other out. On Church-State relationships, while he conceded authority in religious matters to the State, he allowed authority in ecclesiatical matters (appointments, etc.) to the Church, with the proviso that the Church could make over this power to the State. He did not favour a hierarchical Church.
CB 

 
Back
  -  
 
New Search

The deliberate infliction of harm, by authorized agents, on a person, in response to a breach of rules by which, it is claimed, the person is governed, and for which he or she is held responsible. Because of the concentrated coercive power at its disposal, state punishment has been a primary concern of political and legal theory. Here the rules are the laws of the state; the legitimacy of the legal system as a whole is contestable, as is the moral obligation to obey particular laws; and the purposes of the punishment may be variously understood. These purposes are usually identified as deterrence and retribution. Although denunciation, prevention and reform are also mentioned, many theorists would reject these as objects of punishment (rather than possible side-effects or opportunities presented by it). Some accounts of state punishment define it as deliberate infliction of harm on a person who is guilty of breaking the law, in response to that breach, ruling out the possibility that an innocent person may be unjustly punished. Although this is unhelpful, justifiable punishment requires that there be compelling reasons to suppose the person to be punished is guilty; a realistic account of the practice must allow for the possibility of error, even if it hopes to minimize it. The connection between the breach of the rules and the person punished depends upon a conception of responsibility, which is again liable to be controversial, either because of different understandings of the ‘causes’ of a particular individual's behaviour or because of disagreement about the reasonableness of holding X (e.g. a parent, an army officer) responsible for what Y (e.g. a child, a soldier) did.
AR 

 
Back
  -  
 
New Search

The Society of Friends was founded by George Fox in 1650 and nicknamed Quakers because Fox told a judge to ‘tremble in the name of the Lord’ (in other accounts, because Quakers did so themselves). Distinguished in religion for their silent meetings for worship and rejection of ministers and sacraments; and in politics chiefly for:

(1) 
Pacifism
: Fox rejected a request for army service because he ‘lived in the virtue of that life and power that took away the occasion of all wars’. This and other founding statements form the basis of the Society's ‘peace testimony’.
(2) 
Proceeding by consensus rather than by vote in business meetings
. This could be characterized as a unanimity rule.

Through their government of Pennsylvania prior to the 1750s, Quaker doctrines were influential out of proportion to the small size of the Society of Friends.

 
Back
  -  
 
New Search

A quasi non-governmental organization is one created and funded by government, and, therefore, held to account for its expenditure, but given operational independence. The term was invented by Alan Pifer , President of the Carnegie Corporation, to describe such organizations which were appearing in the United States. Subsequently, political scientists, observing the closeness to government of some quangos in their operations, have preferred the term to mean quasi-governmental rather than non-governmental. In the United Kingdom the term has been applied to many forms of arms-length public provision showing a great diversity of purpose, including the BBC, the Welsh Development Agency, and the Commission for Racial Equality. Concern has recently been expressed at the tendency for power to flow from elected public bodies to unelected quangos, derisively dubbed ‘quangocracy’ by some. JBr

 
Back
  -  
 
New Search

The range of mathematical and statistical techniques used to analyse data. In order to test empirical theories and hypotheses, political scientists draw on a wide range of sources, including primarily qualitative data such as documents, unstructured interviews, and participant observation, and primarily quantitative data such as those derived from sample surveys or aggregate statistics such as election results, census materials, or crossnational statistical series.
In order to analyse quantitative data, it is first necessary to describe them, that is, to structure the information and to identify overall patterns. Once these patterns have been established then, secondly, it is important to examine the interrelationships between variables, to see whether they are associated or correlated and if so how strongly. Thirdly, assuming that the researcher has a priori reasons for asserting causal relations between variables, the question then arises of how far changes in the causal or predictor or independent variables can explain changes in the caused or predicted or dependent variables. Finally, if the data are from a sample, the issue arises of how far results can be inferred to be an accurate reflection of the population as a whole. It is to fulfil these four functions—description, association and correlation, explanation, and inference—that political scientists use a range of statistical and mathematical techniques.
The choice of such techniques varies according to a number of considerations, most notably the level of measurement. The lowest level is where phenomena are divided into groups, as in surveys where respondents are classified by socio-economic criteria (e.g. working or middle class, Catholic or Protestant) or partisanship (e.g. Conservative or Labour, Christian Democrat or Social Democrat, Republican or Democrat). Such data are termed nominal- or categorical-level data. The next level is that where phenomena can be ordered in a hierarchy on a ‘more than/less than’ basis; thus, for example, countries may be ranked on whether they are ‘more’ or ‘less’ democratic than, according to appropriate criteria. Such data are termed ordinal-level. The third, and highest, level of measurement is where not only the ordering of phenomena is known, but also the magnitude of differences in terms of a scale; examples might include census data on the social composition of constituencies in terms of percentages in particular groups, or electoral data showing the distribution of the vote between the parties. These are called interval-level data.
For each level of measurement, there is a range of statistics intended to fulfil the functions set out above. Nominal-level data may be described by the mode and the category distributions; association may be measured by statistics such as phi and the contingency coefficient; explanation may be approached via log linear analysis; statistical significance is usually assessed for single variables by looking at the sampling errors of proportions (as with opinion poll predictions) or with association by using a chi-square test. Ordinal-level data are conventionally described by calculating the median and the quartile, octile or decile distribution of the data; the most frequently used measure of association is Spearman's rho; significance is normally assessed by a t-test. Interval-level data are normally described by the mean and the standard deviation, and these can be tested for sampling error; correlation is measured by Pearson's
r
; explanation is catered for by regression analysis; significance is assessed by
T
and
F
tests. In addition, of course, data may be mixed in terms of the levels and there is a range of statistics designed to facilitate this: For example, it may be desired to examine relationships between nominal- and interval-level variables using an analysis of variance test. Finally, where there are many variables of different types, and researchers wish to isolate underlying similarities and differences, there are sophisticated techniques to allow them to do this, including factor, cluster, and discriminant analyses.
Such methods have been widely used by political scientists in a range of contexts, including, for example, the study of arms races, of political stability, of political violence, and of the behaviour of legislators, but by far their most prominent application has been in the area of electoral attitudes and behaviour. Here data are easily quantified, and so such techniques have been extensively used.
While there can be little doubt that quantitative methods have enhanced the study of politics, there have been criticisms of overenthusiasm in their use (quantifying for the sake of it), of equating results obtained with the results of scientific experiments (misapplying the methods of the natural sciences to social data), and overemphasizing numbers at the expense of explanation (the establishment of the existence of a statistically significant correlation or regression coefficient may say little about its meaning). Such criticisms have led some to a more restrained and cautious use of quantitative methods.
ST