Está en la página 1de 43

"The New

Riddle of
Induction."
Chapter III of
Nelson
Goodman
(1983): Fact,
Fiction, and
Forecast.
Cambridge,
MA: Harvard
University
Press, 59-83.

Traducción: Joel Tucídides Madrigal


Bailón, Licenciado en Filosofía, México
Distrito Federal. Universidad Autónoma
Metropolitana Unidad Iztapalapa
(tucidides2000@yahoo.com.mx). Para
educar y sin fines de lucro. Se permite la
reproducción total o parcial citando al
autor original.

Para Angélica, porque su cariño e


inteligencia son mi fuerza en la
adversidad.

Para Carmina Itzel Ruíz Madrigal, porque


su sonrisa es la prueba del buen futuro que
todavía está por venir.
EL NUEVO ENIGMA DE LA INDUCCIÓN
59
NELSON GOODMAN

III III
THE NEW RIDDLE OF INDUCTION EL NUEVO ENIGMA DE LA INDUCCIÓN
POR NELSON GOODMAN 
TRADUCCIÓN: JOEL TUCÍDIDES MADRIGAL BAILÓN

I. The Old Problem of Induction 1. El viejo problema de la inducción

At the close of the preceding lecture, I said that Al cierre de la lectura precedente, Dije que ahora
today I should examine how matters stand examinaría el estado de asuntos respecto al problema
with respect to the problem of induction. In a de la inducción. En concreto pienso que están mal.
word, I think they stand ill. But the real Pero las dificultades reales que nos confrontan hoy, no
difficulties that confront us today are not the son las tradicionales. Lo que es comúnmente pensado
traditional ones. What is commonly thought of como el Problema de la Inducción ha sido resuelto, o
as the Problem of Induction has been solved, más bien, disuelto y encaramos nuevos problemas que
or dissolved; and we face new problems that todavía no son muy bien comprendidos. Para
are not as yet very widely understood. To acercarnos a ellos, tengo que atravesar tan rápido
approach them, I shall have to run as quickly como me sea posible, un terreno que ya nos es muy
as possible over some very familiar ground. familiar.

The problem of the validity of judgments about El problema acerca de la validez de los juicios que
future or unknown cases arises, as Hume tienen que ver con el futuro o casos desconocidos,
pointed out, because such judgments are neither surge (como ya lo señaló Hume) debido a que tales
reports of experience nor logical consequences of juicios no son ni reportes de la experiencia ni
it. Predictions, of course, pertain to what has consecuencias lógicas de ello. Las predicciones, por
not yet been observed. And they cannot be supuesto, atañen a lo que aún no se ha observado, y
logically inferred from what has been observed; no pueden ser lógicamente inferidas de lo que ha sido
for what has happened imposes no logical observado, pues lo que ha sucedido no impone alguna
restrictions on what will happen. Although restricción lógica sobre lo que sucederá. aunque el
Hume's dictum that there are no necessary dictum de Hume de que no hay conexiones necesarias
connections of matters of fact has been sobre las cuestiones de hecho ha sido cambiado a
challenged at times, it has withstood all attacks. veces, este ha resistido todos los ataques y por eso
Indeed, I should be inclined not merely to agree con más razón, me siento inclinado no sólo a estar de
that there are no necessary connections of acuerdo en que no hay conexiones necesarias de
matters of fact, but to ask whether there are any cuestiones de hecho, sino también tentado a
necessary connections at all but that is another investigar si existe alguna conexión necesaria de
story. (1) cualquier clase, pero eso es otra historia. (1)


Tema: Goodman‟s New Riddle of Induction. Artículo inglés de dominio público hallado en red el 15 de mayo del 2011 del vínculo:
google [PDF] THE NEW RIDDLE OF INDUCTION . Esta versión castellana puede descargarse de la página: http://es.scribd.com/
Referencia bibliografía: “The New Riddle of Induction.” Chapter III of Nelson Goodman (1983): Fact, Fiction, and Forecast.
Cambridge, MA: Harvard University Press, 59-83.
Las notas originales (1) se han colocado a final de texto para no distraer la lectura principal y facilitar un manejo bilingüe bajo
el formato Microsoft Word 2007. El acomodo de palabras ha variado ligeramente la paginación: algunos párrafos se han recorrido “al
siguiente inicio de página” respecto a la versión original (vgr. Las líneas últimas de p. 67, se trasladaron a inicio de la 68).
La paginación se maneja así: Arriba en un recuadro negro se pone el número aproximado en el formato original, y abajo la
paginación particular de este documento.
Traducido por Joel Tucídides Madrigal Bailón, Licenciado en Filosofía, México Distrito Federal. Universidad Autónoma
Metropolitana- Unidad Iztapalapa. (tucidides2000@yahoo.com.mx) Para educar y sin fines de lucro. Se permite la reproducción total o
parcial citando al autor original.

Página 1 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
60
NELSON GOODMAN

Hume's answer to the question how La respuesta de Hume a la pregunta del cómo las
predictions are related to past experience is predicciones se relacionan con la experiencia pasada
refreshingly non-cosmic. When an event of one tiene una frescura no - cósmica. Cuando un evento de
kind frequently follows upon an event of another cierto tipo le sigue frecuentemente a otro evento
kind in experience, a habit is formed that leads relativo a otra clase de experiencia; se forma un hábito
the mind, when confronted with a new event of que le indica a la mente (al enfrentarse con un evento
the first kind, to pass to the idea of an event del primer tipo) el paso a la idea de un evento del
of the second kind. The idea of necessary segundo tipo. La idea de conexión necesaria surge del
connection arises from the felt impulse of the impulso encubierto de la mente de realizar esta
mind in making this transition. transición.

Now if we strip this account of all extraneous Ahora si quitamos al asunto todo elemento exterior,
features, the central point is that to the entonces el punto central radica en la pregunta: ¿Por
question "Why one prediction rather than qué elegir una predicción en vez de otra?, Hume
another?", Hume answers that the elect contesta que la predicción electa es una que está de
prediction is one that accords with a past acuerdo con la regularidad pasada, pues esta
regularity, because this regularity has regularidad ha establecido cierto hábito. De ese modo
established a habit. Thus among alternative entre declaraciones alternativas acerca de un
statements about a future moment, one statement momento futuro, un enunciado se distingue por su
is distinguished by its consonance with habit congruencia con el hábito y consecuentemente, por
and thus with regularities observed in the past. las regularidades observadas en el pasado. La
Prediction according to any other alternative is predicción en base a cualquier otra alternativa, es
errant. equivocada.

How satisfactory is this answer? The heaviest ¿Qué tan satisfactoria es esta respuesta? El Criticismo
criticism has taken the righteous position that más estricto ha considerado que la posición que
Hume's account at best pertains only to the Hume, atiende en el mejor de los casos, a las fuentes
source of predictions, not their legitimacy; that de la predicción y no a su legitimidad, esto es:
he sets forth the circumstances under which establece las circunstancias bajo las cuáles hacemos
we make given predictions-and in this sense predicciones --- y esto nos explica por qué hacemos
explains why we make them-but leaves predicciones --- pero ello deja a un lado la pregunta
untouched the question of our license for making por nuestra justificación para realizarlas. Este es el
them. To trace origins, runs the old complaint, problema: Rastrear orígenes o tocar antiguos
is not to establish validity: the real question is problemas, no es establecer validez alguna: La
not why a prediction is in fact made but how it pregunta real no radica en que una predicción sea de
can be justified. Since this seems to point to hecho creada, sino en el cómo puede ser justificada.
the awkward conclusion that the greatest of Esto da la impresión de apuntar hacia la conclusión
modern philosophers completely missed the point embarazosa, de que los más grandes filósofos
of his own problem, the idea has developed that modernos perdieron de vista el auténtico problema.
he did not really take his solution very seriously, La idea que se ha desarrollado alternativamente es
but regarded the main problem as unsolved and que Hume realmente no tomó en serio su propia
perhaps as insoluble. Thus we come to speak of solución, sino que estimó el problema principal como
'Hume's problem' as though he propounded it as algo no resuelto y quizás insoluble. Así llegamos a
a question without answer. hablar del Problema de Hume como algo que él
construye a la manera de una pregunta sin respuesta.

Página 2 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 61

All this seems to me quite wrong. I think Hume Todo esto me parece muy mal. Pienso que Hume
grasped the central question and considered captó la pregunta central y consideró su respuesta
his answer to be passably effective. And I como algo básicamente efectivo. Opino que su
think his answer is reasonable and relevant, respuesta es razonable y pertinente, aun si no es
even if it is not entirely satisfactory. I shall enteramente satisfactoria. Por ahora solamente quiero
explain presently. At the moment, I merely want apuntar una protesta en contra de la noción común
to record a protest against the prevalent notion que se tiene sobre la justificación de la inducción,
that the problem of justifying induction, when it cuando acontece que ella, se disocia de la cuestión
is so sharply dissociated from the problem of descriptiva del cómo toma lugar la inducción, y que
describing how induction takes place, can fairly puede ser llamada: El Problema de Hume.
be called Hume's problem.

I suppose that the problem of justifying Supongo que el problema sobre la justificación de la
induction has called forth as much fruitless Inducción ha dado pie a muchos debates sin sentido,
discussion as has any halfway respectable como suele pasar con aquellos problemas
problem of modern philosophy. The typical medianamente respetables de la filosofía moderna. El
writer begins by insisting that some way of escritor típico empieza insistiendo que ciertas formas
justifying predictions must be found; proceeds to de justificación de las predicciones deben ser
argue that for this purpose we need some aclaradas. Procede a sostener que con este propósito,
resounding universal law of the Uniformity of necesitamos cierta ley universal sobre la uniformidad
Nature, and then inquires how this universal de la naturaleza, y entonces indaga cómo este
principle itself can be justified. At this point, principio universal mismo, puede ser justificado. En
if he is tired, he concludes that the principle este punto, si él está cansado, entonces concluye que
must be accepted as an indispensable el principio debe ser aceptado como una suposición
assumption; or if he is energetic and ingenious, indispensable; o si él es drástico e ingenioso, entonces
he goes on to devise some subtle justification for procede a idear alguna justificación sutil para ella. Tal
it. Such an invention, however, seldom improvisación, sin embargo, rara vez satisface a los
satisfies anyone else; and the easier course of demás; y la facilidad de aceptar una suposición no
accepting an unsubstantiated and even dubious confirmada y aun de dudosa condición resulta más
assumption much more sweeping than any actual forzada que cualquiera de las predicciones reales que
predictions we make seems an odd and expensive hacemos, y parece una forma extraña y cara de
way of justifying them. justificarlas.

Página 3 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 62
2. Dissolution of the Old Problem 2. Disolución del viejo problema.

Understandably, then, more critical thinkers Comprensiblemente, luego, varios pensadores críticos
have suspected that there might be something han sospechado que podría haber algo fuera de lugar
awry with the problem we are trying to solve. con el problema que tratamos de resolver. Pensando
Come to think of it, what precisely would bien esto ¿Qué podría constituir con precisión la
constitute the justification we seek? If the justificación que buscamos? Si el problema es explicar
problem is to explain how we know that certain cómo sabemos que ciertas predicciones resultarán ser
predictions will turn out to be correct, the correctas, entonces la respuesta adecuada es que no
sufficient answer is that we don't how any such hay forma de saber tal cosa. Si el problema es
thing. If the problem is to find some way of encontrar algún antecedente distintivo entre
distinguishing antecedently between true and predicciones verdaderas y falsas, entonces nos
false predictions, we are asking for prevision encontramos a la búsqueda de previsión más que de
rather than for philosophical explanation. Nor una explicación filosófica. Tampoco ayuda mucho al
does it help matters much to say that we are asunto decir que solamente estamos tratando de
merely trying to show that or why certain mostrar el qué o el por qué, ciertas predicciones son
predictions are probable. Often it is said that probables. Se dice que mientras no podamos decir por
while we cannot tell in advance whether a adelantado si la predicción de un juego de dados será
prediction concerning a given throw of it die is verdadera, entonces todavía estaremos estableciendo
true, we can decide whether the prediction is a si la predicción es solamente algo probable. No
probable one. But if this means determining how obstante, si esto implica definir cómo es que la
the prediction is related to actual frequency predicción se relaciona con “la distribución real de
distributions of future throws of the die, surely frecuencia de lanzamientos futuros del dado”,
there is no way of knowing or proving this in entonces diremos seguramente que no hay forma de
advance. On the other hand, if the judgment that saber o probar esto por adelantado. Por otra parte, si
the prediction is probable has nothing to do with el juicio de que la predicción es probable, no tiene
subsequent occurrences, then the question nada que ver con los subsiguientes casos, entonces la
remains in what sense a probable prediction is pregunta será: ¿En qué sentido una predicción
any better justified than an improbable one. probable, estará mejor justificada que alguna
improbable?

Now obviously the genuine problem cannot be Ahora vemos obviamente que el verdadero problema
one of attaining unattainable knowledge or of no puede ser el de lograr un conocimiento inalcanzable
accounting for knowledge that we do not in fact o el de dar cuenta de un conocimiento que en realidad
have. A better understanding of our problem can no tenemos. Una mejor comprensión de nuestro
be gained by looking for a moment at what is problema puede ser obtenida contrastando las cosas
involved in justifying non-inductive inferences. con otro momento donde se halle envuelta la
How do we justify a deduction? Plainly, by justificación de inferencias no - inductivas. Pensando
showing that it conforms to the general rules of así ¿Cómo justificamos una deducción?
deductive inference. An argument that so Explícitamente, demostrando que obedece a las reglas
conforms is justified or valid, even if its generales de la inferencia deductiva. Un argumento
conclusion happens to be false. An argument that que se adapta, es justificado o válido, aun si su
violates a rule is fallacious even if its conclusion conclusión llega a ser falsa. Un argumento que
happens to be true. To justify a deductive contraviene una ley, es falaz aun si su deducción
conclusion therefore requires no knowledge of the acierta a ser verdadero. Justificar una conclusión
facts it pertains to. Moreover, when a deductive deductiva por consiguiente no precisa conocimiento
argument has been shown to conform to the rules del hecho al que pertenece. Aún más, cuando un

Página 4 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
63
NELSON GOODMAN

of logical inference, we usually consider it argumento deductivo ha sido expresado para obedecer
justified without going on to ask what justifies the las reglas de inferencia lógica, usualmente lo
rules. Analogously, the basic task in justifying an consideramos justificado sin llegar a preguntar qué
inductive inference is to show that it conforms to justifica a las reglas. Análogamente, la tarea básica
the general rules of induction. Once we have de justificar una inferencia inductiva, es demostrar que
recognized this, we have gone a long way obedece a las reglas generales de la inducción. Una
towards clarifying our problem. vez que hemos reconocido esto, entonces ya
recorrimos un gran camino para aclarar nuestro
problema.

Yet, of course, the rules themselves must Todavía así, por supuesto, las reglas por sí mismas
eventually be justified. The validity of a tienen que estar eventualmente justificadas. La validez
deduction depends not upon conformity to any de una deducción depende no de la conformidad o
purely arbitrary rules we may contrive, but upon coherencia que tenga con alguna regla arbitraria que
conformity to valid rules. When we speak of the nosotros podamos construir, si no de la conformidad o
rules of inference we mean the valid rules or coherencia que tenga con reglas válidas. Cuando
better, some valid rules, since there may be hablamos de las reglas de la inferencia nos queremos
alternative sets of equally valid rules. But how is referir a reglas válidas, o mejor dicho, a algunas reglas
the validity of rules to be determined? Here again válidas, ya que podría haber un escenario alternativo
we encounter philosophers who insist that these para reglas igualmente válidas ¿Pero cómo es que las
rules follow from some self-evident axiom, and reglas puede ser determinadas? Aquí encontramos de
others who try to show that the rules are grounded nuevo a los filósofos que insisten en el tema de que
in the very nature of the human mind. I think the estas reglas se siguen de cierto axioma auto-evidente,
answer lies much nearer the surface. Principles of y otros que tratan de mostrar que las reglas están
deductive inference are justified by their fundamentadas en la muy particular naturaleza de la
conformity with accepted deductive practice. mente humana. Opino que la respuesta descansa muy
Their validity depends upon accordance with the cerca de la superficie. Los principios de la inferencia
particular deductive inferences we actually make deductiva están justificados por su conformidad con
and sanction. If a rule yields inacceptable una práctica deductiva aceptada. Su validez depende
inferences, we drop it as invalid. Justification of de su coherencia con las inferencias deductivas
general rules thus derives from judgments particulares que realmente toman efecto y sanción. Si
rejecting or accepting particular deductive una regla produce inferencias inaceptables, nosotros la
inferences. descartamos como algo inválido. Entonces, la
justificación de las reglas generales se derivan de
juicios que rechazan o aceptan inferencias deductivas
particulares.

Página 5 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 64

This looks flagrantly circular. I have said that Esto nos aparece evidentemente circular. He dicho
deductive inferences are justified by their que las inferencias deductivas están justificadas por
conformity to valid general rules, and that general su conformidad a reglas generales válidas, y que las
rules are justified by their conformity to valid reglas generales están justificadas por su conformidad
inferences. But this circle is a virtuous one. The a inferencias válidas. Pero este es un círculo virtuoso.
point is that rules and particular inferences alike El punto es que las reglas y las inferencias
are justified by being brought into agreement with particulares semejantes entre si, se justifican cuando
each other. A rule is amended if it yields an llegan a un acuerdo unas con otras. Una regla es
inference we are unwilling to accept; an inference modificada si da lugar a una inferencia que nos sea
is rejected if it violates a rule we are unwilling to difícil de aceptar; una inferencia es rechazada si viola
amend. The process of justification is the delicate alguna regla que nos sea difícil de modificar. El
one of making mutual adjustments between rules proceso de justificación es uno muy delicado en el que
and accepted inferences; and in the agreement se realizan ajustes entre reglas e inferencias
achieved lies the only justification needed for aceptadas, y en el acuerdo alcanzado descansa la
either. única justificación necesaria a cada una.

All this applies equally well to induction. An Todo esto aplica por igual a la inducción. Una
inductive inference, too, is justified by conformity inferencia inductiva también es justificada por su
to general rules, and a general rule by conformity conformidad a inferencias inductivas aceptadas. Las
to accepted inductive inferences. Predictions are predicciones son justificadas si se apegan a cánones
justified if they conform to valid canons of válidos de la inducción; y los cánones son válidos si se
induction; and the canons are valid if they corresponden con precisión a una práctica inductiva
accurately codify accepted inductive practice. aceptada.

A result of such analysis is that we can stop El resultado de tal análisis es que podemos dejar de
plaguing ourselves with certain spurious atormentarnos con algunas preguntas falsas acerca de
questions about induction. We no longer demand la inducción. No pedimos más explicaciones para
an explanation for guarantees that we do not garantizar lo que no tenemos, o bien para obtener
have, or seek keys to knowledge that we cannot ciertas claves de un conocimiento que no podemos
obtain. It dawns upon us that the traditional smug tener. Nos resulta claro que la soberbia insistencia
insistence upon a hard-and-fast line between tradicional en una delimitación pura, clara y rápida
justifying induction and describing ordinary respecto a la justificación de la inducción y la
inductive practice distorts the problem. And we descripción de la práctica ordinaria inductiva, ha
owe belated apologies to Hume. For in dealing distorsionado el problema. Por esto le debemos unas
with the question how normally accepted disculpas a Hume. Esto porque al encargarse de la
inductive judgments are made, he was in fact pregunta que tan comúnmente aceptaba, sobre el
dealing with the question of inductive validity. (2) cómo son hechos los juicios inductivos; él estaba de
The validity of a prediction consisted for him in hecho, ocupándose de la pregunta de la validez
its arising from habit, and thus in its exemplifying inductiva. (2) La validez de una predicción se
some past regularity. His answer was incomplete originaba para Hume, en un arraigo de los hábitos, y
and perhaps not entirely correct; but it was not por ello en cierto ejemplo de una regularidad pasada.
beside the point. The problem of induction is not Su respuesta pudiera estar incompleta y quizás no
a problem of demonstration but a problem of enteramente en lo correcto; pero ello no venía al caso.
defining the difference between valid and invalid El problema de la inducción no es un problema de
predictions. demostración sino un problema de definir la diferencia
entre predicciones válidas e inválidas.

Página 6 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 65

This clears the air but leaves a lot to be done. As Esto despeja el ambiente, pero nos deja con mucho
principles of deductive inference, we have the qué hacer. Como principios de la inferencia deductiva,
familiar and highly developed laws of logic; but tenemos las familiares y altamente desarrolladas leyes
there are available no such precisely stated and de la lógica pero no se establecen precisamente como
well-recognized principles of inductive inference. reconocidos principios de la inferencia inductiva. Los
Mill's canons hardly rank with Aristotle's rules of cánones de Mill difícilmente están al mismo nivel que
the syllogism, let alone with Principia las reglas del silogismo de Aristóteles, y mucho menos
Mathematics. Elaborate and valuable treatises on al nivel del Principia Matemática. Los tratados más
probability usually leave certain fundamental elaborados y valiosos sobre la probabilidad
questions untouched. Only in very recent years usualmente dejan ciertas preguntas fundamentales sin
has there been any explicit and systematic work tocar. Sólo en años recientes ha llegado a existir algún
upon what I call the constructive task of trabajo explícito y sistemático sobre lo que llamo la
confirmation theory. tarea constructiva de la teoría de la confirmación.

3. The Constructive Task of Confirmation Theory 3. La Tarea Constructiva de la Teoría de la


Confirmación.

The task of formulating rules that define the La tarea de formular reglas que definan las diferencias
difference between valid and invalid inductive entre inferencias inductivas válidas e inválidas es
inferences is much like the task of defining any parecida a la tarea de definir algún término a través de
term. with an established usage. If we set out to un uso establecido. Si intentamos definir el término
define the term "tree", we try to compose out of "árbol" trataremos de arreglar en base a palabras ya
already understood words an expression that will conocidas una expresión que aplicará a los objetos
apply to the familiar objects that standard usage familiares de tal manera que el uso estándar se
calls trees, and that will not apply to objects that denomine: "árboles", y que no se aplicará a los objetos
standard usage refuses to call trees. A proposal a los que el uso estándar rechace llamar "árboles".
that plainly violates either condition is rejected; Una propuesta que explícitamente viole cualquiera de
while a definition that meets these tests may be estas condiciones, es rechazada; Mientras que una
adopted and used to decide cases that are not definición que reúna estos requisitos puede ser
already settled by actual usage. Thus the interplay adoptada y usada para decidir los casos que todavía
we observed between rules of induction and no son determinados por el uso real. De esta forma, la
particular inductive inferences is simply an coyuntura que observamos entre reglas de inducción e
instance of this characteristic dual adjustment inferencias inductivas particulares es simplemente una
between definition and usage, whereby the usage instancia de este característico ajuste entre la
informs the definition, which in turn guides definición y el uso, a través del cual el uso instruye a la
extension of the usage. definición, que a su vez guía a la extensión de uso.

Página 7 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 66

Of course this adjustment is a more complex Por supuesto este ajuste es un asunto más complejo
matter than I have indicated. Sometimes, in the de lo que he indicado. Algunas veces, en interés de la
interest of convenience or theoretical utility, we conveniencia o la utilidad teórica, deliberadamente
deliberately permit a definition to run counter to permitimos a una definición ir en contra de mandatos
clear mandates of common usage. We accept a de claro uso común. Aceptamos una definición de
definition of "fish" that excludes whales. "pez" que excluye ballenas. De modo semejante
Similarly we may decide to deny the term "valid podemos resolver negar el término "Inducción válida"
induction" to some inductive inferences that are a algunas inferencias inductivas que son comúnmente
commonly considered valid, or apply the term to consideradas válidas, o aplicar el término a otras
others not usually so considered. A definition inferencias que usualmente no están considerados de
may modify as well as extend ordinary usage. (3) esta manera. Una definición puede modificar así como
también extender, el uso ordinario. (3)

Some pioneer work on the problem of defining Cierto trabajo pionero sobre el problema de "la
confirmation or valid induction has been done by confirmación de la definición" ó inducción válida, ha
Professor Hempel. (4). Let me remind you briefly sido realizado por el profesor Hempel. (4) Permítame
of a few of his results. Just as deductive logic is mostrar brevemente unos pocos de sus resultados.
concerned primarily with a relation between Así como la lógica deductiva está preocupada
statements-namely the consequence relation-that primordialmente por la relación entre declaraciones ---
is independent of their truth or falsity, so particularmente las consecuencias de estas relaciones
inductive logic as Hempel conceives it is --- y ello es independiente de su verdad o su
concerned primarily with a comparable relation falsedad, así la lógica inductiva, como Hempel la
of confirmation between statements. Thus the concibe, está preocupada primordialmente por la
problem is to define the relation that obtains relación comparable de confirmación entre
between any statement S1 and another Sz if and declaraciones. De esta forma, el problema es definir la
only if S1 may properly be said to confirm SZ in relación que se obtiene entre cualquier declaración
any degree. “S1” y otra “Sz”, si y sólo si, “S1” efectivamente
confirma a “SZ” en cualquier grado.

Página 8 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 67

With the question so stated, the first step seems Así planteada la cuestión, el primer paso luce obvio
obvious. Does not induction proceed in just the ¿Acaso no sucede que la inducción procede tan sólo
opposite direction from deduction? Surely some de la dirección opuesta de la deducción?
of the evidence statements that inductively Seguramente algunas de las declaratorias de
support a general hypothesis are consequences of evidencia que soportan inductivamente una hipótesis
it. Since the consequence relation is already well son consecuencias de ella. Ya que la relación de
defined by deductive logic, will we not be on consecuencia está de hecho bien definida por lógica
firm ground in saying that confirmation embraces deductiva, ¿estaremos en "terrenos firmes" al decir
the converse relation? The laws of deduction in que nuestra confirmación implica la relación inversa?
reverse will then be among the laws of induction. La leyes de la deducción a la inversa, se encontrarían
entonces entre las leyes de la inducción.

Let's see where this leads us. We naturally Veamos dónde nos dirige esto. Naturalmente
assume further that whatever confirms a given asumimos además que lo que fuere que confirme una
statement confirms also whatever follows from declaración dada, confirma también lo que esté detrás
that statement. (5) But if we combine this de esa declaración. (5) Pero Si nosotros combinamos
assumption with our proposed principle, we get esta suposición con nuestro principio declarado,
the embarrassing result that every statement obtenemos el molesto resultado de que cada
confirms every other. Surprising as it may be that declaración confirma a todas los demás. Es
such innocent beginnings lead to such an sorprendente el como puede ser que tales orígenes
intolerable conclusion, the proof is very easy. tan inocentes nos conduzcan a conclusiones tan
Start with any statement SI. It is a consequence intolerables, la prueba es muy simple. Comience con
of, and so by our present criterion confirms, the cualquier declaración S1 que sea una consecuencia
conjunction of S1 and any statement what so ever (y por lo tanto en base a nuestro criterio, también “la
call it S2. But the confirmed conjunction, Sl*S2, confirmación”), de la conjunción de S1 y cualquier
of course has S2 as a consequence. Thus every otra declaración a la que llamaremos en todo
statement confirms all statements. momento S2. Tendremos que la conjunción
confirmada, S1 x S2, incluirá por supuesto a S2 como
consecuencia. Luego entonces, cada declaración
confirmará todas las demás declaraciones.

Página 9 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 68

The fault lies in careless formulation of our first El defecto yace en una formulación confusa de nuestra
proposal. While some statements that confirm a primera propuesta. Mientras que algunos enunciados
general hypothesis are consequences of it, not all que confirman una hipótesis general, son
its consequences confirm it. This may not be consecuencias de ella, no todas sus consecuencias la
immediately evident; for indeed we do in some confirman. Esto no siempre puede ser tan evidente,
sense furnish support for a statement when we pues en efecto en algún sentido proveemos soporte
establish one of its consequences. We settle one para una declaración cuando establecemos una de
of the questions about it. Consider the sus consecuencias. Sobre el tema estableceremos
heterogeneous conjunction: 8497 is a prime una pregunta. Considere esta conjunción
number and the other side of the moon is flat and heterogénea: "8497 es un número primo", "El otro
Elizabeth the First was crowned on a Tuesday. lado de la luna es plano" y "Elizabeth Primera fue
coronada en un martes".

To show that any one of the three component Demostrar que cualesquiera de las tres declaraciones
statements is true is to support the conjunction by componentes es cierta, equivale a sostener la
reducing the net undetermined claim. But support conjunción reduciendo la red de juicios
(6) of this kind is not confirmation; for indeterminados. Pero una prueba de este tipo (6) no
establishment of one component endows the es una confirmación, pues la estabilidad de un
whole statement with no credibility that is componente no vincula a la declaración entera con
transmitted to other component statements. una credibilidad transmitida a otras declaraciones
Confirmation of a hypothesis occurs only when componentes. La confirmación de una hipótesis ocurre
an instance imparts to the hypothesis some sólo cuando una instancia imparte para la hipótesis
credibility that is conveyed to other instances. alguna credibilidad, que es comunicada a otras
Appraisal of hypothesis, indeed, is incidental to instancias. El avalúo de la hipótesis, ciertamente, es
prediction, to the judgment of new cases on the incidental para la predicción, es decir, para el juicio de
basis of old ones. casos nuevos con base en viejos casos.

Página 10 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 69

Our formula thus needs tightening. This is readily Entonces nuestra fórmula necesita mejores bases.
accomplished, as Hempel points out, if we Esto se concluye fácilmente, como Hempel lo apunta,
observe that a hypothesis is genuinely confirmed si observamos que una hipótesis está genuinamente
only by a statement that is an instance of it in the confirmada sólo por una declaración que a su vez es
special sense of entailing not the hypothesis itself una instancia de ella, en el sentido especial, de no
but its relativization or restriction to the class of conllevar la hipótesis misma sino su relativización o su
entities mentioned by that statement. The restricción a la clase de entidades mencionadas por
relativization of a general hypothesis to a class esa declaración. La relativización de una hipótesis
results from restricting the range of its universal general para una clase resulta de restringir el rango de
and existential quantifiers to the members of that sus cuantificadores universales y existenciales para los
class. Less technically, what the hypothesis says miembros de esa clase. Menos técnicamente, lo que la
of all things the evidence statement says of one hipótesis dice de todas las cosas, de las que habla un
thing (or of one pair or other wad of things). This "enunciado o declaratoria de evidencia"; lo dice
obviously covers the confirmation of the también de una cosa en particular (ó también de un
conductivity of all copper by the conductivity of a par u cualquier otro conjunto de elementos). Esto
given piece; and it excludes confirmation of our obviamente se relaciona con la confirmación de "la
heterogeneous conjunction by any of its conductividad del cobre a través de la conductividad
components. ,4nd, when taken together with the de un trozo dado”; y excluye la confirmación de
principle that what confirms a statement confirms nuestra “conjunción heterogénea” por cualquiera de
all its consequences, this criterion does not yield sus componentes. Luego entonces, cuándo esto es
the untoward conclusion that every statement tomado conjuntamente con el principio de que "una
confirms every other. declaración confirma todas sus consecuencias", este
criterio no produce la conclusión inversa de que cada
declaración "confirma en todas las otras".

New difficulties promptly appear from other Sin embargo, nuevas dificultades aparecen
directions, however. One is the infamous paradox inmediatamente desde otras direcciones. Una es la
of .the ravens. The statement that a given object, infame paradoja de los cuervos. La afirmación de que
say this piece of paper, is neither black nor a un objeto dado, por decir esta pieza de papel, "no es ni
raven confirms the hypothesis that all non-black negra ni tampoco un cuervo", confirma la hipótesis de
things are non-ravens. But this hypothesis is que toda cosa "no negra" tampoco es "un cuervo".
logically equivalent to the hypothesis that all Pero esta hipótesis es lógicamente equivalente a
ravens are black. Hence we arrive at the aquella hipótesis de que "Todos los cuervos son
unexpected conclusion that the statement that a negros". Así llegamos a la inesperada conclusión
given object is neither black nor a raven confirms donde la afirmación de que un objeto dado no es "ni
the hypothesis that all ravens are black. The negro ni un cuervo", confirma también la hipótesis de
prospect of being able to investigate que "Todos los cuervos son negros". La esperanza de
ornithological theories without going out in the poder investigar teorías ontológicas "sin mojarse bajo
rain is so attractive that we know there must be a la lluvia" es tan atractiva que sabemos que puede
catch in it. The trouble this time, however, lies haber algún truco en ella. El problema esta vez, sin
not in faulty definition, but in tacit and illicit embargo, no recae sobre una definición deficiente,
reference to evidence not stated in our example. sino en la referencia tácita e ilícita a una evidencia no
Taken by itself, the statement that the given indicada en nuestro ejemplo. Tomada por si sola, la
object is neither black nor a raven confirms the declaración que el objeto dado es "ni negro ni un
hypothesis that everything that is not a raven is cuervo" confirma que la hipótesis de que "todo lo que

Página 11 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
70
NELSON GOODMAN

not black as well as the hypothesis that everything no sea un cuervo no es negro", así como también la
that is not black is not a raven. We tend to ignore hipótesis que "todo lo que no sea negro no es un
the former hypothesis because we know it to be cuervo". Entonces observamos que tendemos a
false from abundant other evidence-from all the ignorar la hipótesis anterior porque la sabemos falsa --
familiar things that are not ravens but are black. a partir de distintas y abundantes pruebas de todas las
But we are required to assume that no such cosas familiares que "no son cuervos ni son negras".
evidence is available. Under this circumstance, Pero en este caso, estamos obligados a asumir que
even a much stronger hypothesis is also obviously nada de tal evidencia está disponible. Bajo esta
confirmed: that nothing is either black or a raven. circunstancia incluso una hipótesis mucho más fuerte
In light of this confirmation of the hypothesis that y extravagante, también llegar a ser confirmada, por
there are no ravens, it is no longer surprising that ejemplo que: “ nada puede ser ni negro, ni tampoco un
under the artificial restrictions of the example, the cuervo”. En vista de esta confirmación de la hipótesis
hypothesis that all ravens are black is also de que "no hay cuervos", ya no es sorprendente que
confirmed. And the prospects for indoor bajo las restricciones artificiales del ejemplo, la
ornithology vanish when we notice that under hipótesis de que "todos los cuervos son negros" sea
these same conditions, the contrary hypothesis también confirmada. Y los prospectos para la
that no ravens are black is equally well ornitología interior desaparecen cuando advertimos
confirmed. (7) que bajo estas mismas condiciones, la hipótesis
inversa a: "ninguno de los cuervos son negros" está
igualmente bien confirmada. (7)

On the other hand, our definition does err in not Por otra parte, nuestra definición se equivoca al no
forcing us to take into account all the stated forzarnos a tomar en cuenta toda la evidencia
evidence. The unhappy results are readily establecida. Los desafortunados resultados son
illustrated. If two compatible evidence statements realmente ilustrativos. Si dos enunciaos de evidencia
confirm two hypotheses, then naturally the confirman dos hipótesis, entonces naturalmente la
conjunction of the evidence statements should conjunción de los enunciados de evidencia deberían
confirm the conjunction of the hypotheses (8) confirmar la conjunción de las hipótesis. (8) Suponga
Suppose our evidence consists of the statements que nuestra evidencia consiste en los enunciados: a)
EI saying that a given thing b is black, and E2 E1 el cual afirma que una cosa dada "b" es negra, y b)
saying that a second thing c is not black. By our E2 el cual dice un segundo elemento "c" no es blanco.
present definition, EI confirms the hypothesis that En razón de nuestra presente definición, E1 confirma
everything is black, and E2 the hypothesis that la hipótesis de que cualquier cosa es negra, y E2 la
everything is non-black. The conjunction of these hipótesis de que cualquier cosa no es negra. La
perfectly compatible evidence statements will conjunción de estos enunciados de evidencia
then confirm the self-contradictory hypothesis perfectamente compatibles confirmarán entonces la
that everything is both black and nonblack. auto-contradictoria hipótesis de que cualquier cosa
Simple as this anomaly is, it requires drastic puede ser ambas: "negra" y "no-negra". La simpleza
modification of our definition. de esta anomalía nos exige una modificación drástica
de nuestra definición.

Página 12 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN

71

What given evidence confirms is not what we Lo que confirma una cierta evidencia, no es tanto a lo
arrive at by generalizing from separate items of it, que llegamos cuando partimos de la generalización de
but -roughly speaking- what we arrive at by sus elementos separados, sino --- hablando a grandes
generalizing from the total stated evidence. The rasgos --- a lo que llegamos con la generalización de la
central idea for an improved definition is that, evidencia total establecida. La idea central para una
within certain limitations, what is asserted to be definición mejorada es que, dentro de ciertas
true for the narrow universe of the evidence limitaciones, lo que es afirmado como cierto para el
statements is confirmed for the whole universe of pequeño universo de las declaraciones de evidencia,
discourse. Thus if our evidence is El and Es es confirmado para el universo entero del discurso.
neither the hypothesis that all things are black nor Entonces si nuestra evidencia es El y E2; ni la
the hypothesis that all things are non-black is hipótesis de que todas las cosas son negras, ni la
confirmed; for neither is true for the evidence- hipótesis de que ninguna cosa lo es, llegará a
universe consisting of b and c. Of course, much confirmarse; pues ninguno de las dos es verdadera
more careful formulation is needed, since some debido al universo de evidencia consistente en b y c.
statements that are true of the evidence-universe- Por supuesto, una formulación mucho más cuidadosa
such as that there is only one black thing -are es necesaria, ya que algunas declaraciones del
obviously not confirmed for the whole universe. universo de la evidencia que son ciertas --- tal como
These matters are taken care of by the studied que existe sólo una cosa negra --- definitivamente no
formal definition that Hempel develops on this están confirmadas para el universo entero. Estas
basis; but we cannot and need not go into further situaciones son materia de reflexión en el estudio
detail here. formal de la definición que Hempel desarrolla sobre
estas bases; pero no podemos y no necesitamos
entrar en más detalle aquí.

No one supposes that the task of confirmation- Nadie supone que la tarea de teoría de confirmación
theory has been completed. But the few steps I haya sido completada. Pero los pocos pasos que yo he
have reviewed chosen --- partly for their bearing revisado --- seleccionados en parte por su relación
on what is to follow -- show how things move respecto lo que haya de seguir - nos muestra el cómo
along once the problem of definition displaces the las cosas siguen adelante una vez que el problema de
problem of justification. Important and long la definición, desplaza al problema de la justificación.
unnoticed questions are brought. to .light and Las preguntas importantes y bastante inadvertidas son
answered; and we are encouraged to expect that traídas para aclararlas y contestarlas; Y nosotros
the many remaining questions will in time yield estamos animados para pretender que las muchas
to similar treatment. preguntas restantes con el tiempo se avendrán a un
tratamiento similar.

But our satisfaction is short lived. New and Pero nuestra satisfacción es de corta duración. Nuevos
serious trouble begins to appear. y serios problemas comienzan a aparecer.

Página 13 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 72

4. The New Riddle of Induction 4. El nuevo enigma de la inducción

Confirmation of a hypothesis by an instance La confirmación de una hipótesis en un momento dado


depends rather heavily upon features of the depende en mayor medida, más bien, de las
hypothesis other than its syntactical form. That a características de la hipótesis más que de su forma
given piece of copper conducts electricity sintáctica. Que una pieza de cobre dada conduzca
increases the credibility of statements asserting electricidad, incrementa la credibilidad de enunciados
that other pieces of copper conduct electricity, asertivos de que "Otras piezas de cobre también
and thus confirms the hypothesis that all copper conducen electricidad" y de esta manera se confirma la
conducts electricity. But the fact that a given 'man hipótesis de que "Todo cobre conduce electricidad".
now in this room is a third son does not increase Pero el hecho de que en este instante, cierto hombre
the credibility of statements asserting that other de esta sala "sea un hijo menor", no incrementa la
men now in this room are third sons, and so does credibilidad de enunciados asertivos diciendo que
not confirm the hypothesis that all men now in algún otro hombre de la sala, en este instante, también
this room are third sons. Yet in both cases our sea "un hijo menor", y consecuentemente, no confirma
hypothesis is a generalization of the evidence la hipótesis de que "todo hombre, ahora y en esta sala,
statement. The difference is that in the former sea el hijo menor”. Todavía en ambos casos, nuestra
case the hypothesis is a law like statement; while hipótesis es una generalización de un enunciado de
in the latter case, the hypothesis is a merely evidencia. La diferencia es que, en el primer caso
contingent or accidental generality. Only a presentado, la hipótesis es una afirmación
statement that is law like-regardless of its truth or "legaliforme"  (establecida con cierta formalidad),
falsity or its scientific importance-is capable of mientras que en el caso posterior, la hipótesis es
receiving confirmation from an instance of it; meramente una generalización contingente o
accidental statements are not. Plainly, then, we accidental. Solamente una afirmación que es
must look for a way of distinguishing law like "legaliforme" --- independientemente de su verdad, su
from accidental statements. falsedad o su importancia científica --- es capaz de
recibir confirmación de una instancia de sí misma,
pues los enunciados accidentales no pueden hacerlo.
Evidentemente, debemos buscar una forma para
distinguir los enunciados accidentales de los
"legaliformes".


[Nota de Traducción N.T. ] Legaliforme es un término relativo al tema de la justificación en las teorías
científicas. Lo legaliforme implica cierta legitimidad en una expresión científica. Mario Bunge comenta:
„Fórmula legalforme‟ (o „enunciado nomológíco‟) designa una proposición o [unción proposicional que se
supone normalmente que describe una ley o una parte de una ley (pauta nómica). Una fórmula legaliforme es
un objeto conceptual, a saber, una hipótesis cientitica que satisface ciertos requisitos de generalidad,
corroboración y sisternaticidad (cf. scc. 6.6). No hay necesidad de decir que las leyes que hallarnos en los
textos cienil ficos son en su mayor parte enunciados legaliformes.” [Mario Bunge, "La investigación
científica: su estrategia y su filosofía", p. 300]

Página 14 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 73

So long as what seems to be needed is merely a Mientras que nuestra condición de necesidad sea
way of excluding a few odd and unwanted cases meramente una forma de excluir algunos casos
that are inadvertently admitted by our definition extraños y no deseados, cuyas presencias sean
of confirmation, the problem may not seem very inadvertidamente admitidas por nuestra definición de
hard or very pressing. We fully expect that minor "confirmación", el problema no será muy difícil ni muy
defects will be found in our definition and that the apremiante. Solemos pretender que los defectos
necessary refinements will have to be worked out menores sean encontrados en nuestra definición y que
patiently one after another. But some further los refinamientos necesarios tengan que estar
examples will show that our present difficulty is trabajados pacientemente uno tras otro. Pero algunos
of a much graver kind. ejemplos de más adelante mostrarán que nuestra
dificultad presente es de una clase más profunda.

Suppose that all emeralds examined before a Suponga que todas las esmeraldas examinadas antes
certain time t are green. (9) At time t, then, our de cierto tiempo "t" son verdes. (9) Entonces, en el
observations support the hypothesis that all tiempo "t" nuestras observaciones soportan la
emeralds are green; and this is in accord with our hipótesis de que todas las esmeraldas son verdes; y
definition of confirmation. Our evidence ello está de acuerdo con nuestra definición de
statements assert that emerald a is green, that confirmación. Nuestros enunciados de evidencia
emerald b is green, and so on; and each confirms afirman que cierta esmeralda "a" es verde, que cierta
the general hypothesis that all emeralds are green. esmeralda "b" es verde y así sucesivamente, y cada
So far, so good. caso confirma la hipótesis general de que todas las
esmeraldas son verdes. Hasta ahora, todo bien.

Página 15 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN
74

Now let me introduce another predicate less Ahora déjeme introducir otro predicado menos familiar
familiar than "green". It is the predicate "grue" que "verde". Es el predicado "verdul" (o “grue” en
and it applies to all things examined before t just inglés) y eso se aplica a todas las cosas examinadas
in case they are green but to other things just in antes de “t”, solo en este lapso son verdes, pero en
case they are blue. Then at time t we have, for otros casos serán azules. Luego en el tiempo "t"
each evidence statement asserting that a given tendremos que para cada afirmación de evidencia
emerald is green, a parallel evidence statement afirmando que una esmeralda dada es verde, habrá
asserting that that emerald is grue. And the una declaración paralela de evidencia aseverando que
statements that emerald a is grue, that emerald b esa esmeralda es "verdul". Y las declaraciones que la
is grue, and so on, will each confirm the general esmeralda "a" es verdul, que la esmeralda "b" es
hypothesis that all emeralds are grue. Thus verdul, etcétera, confirmarán cada una la hipótesis
according to our definition, the prediction that all general de que todas las esmeraldas son "verdul". Así
emeralds subsequently examined will be green según nuestra definición, la predicción de que todas
and the prediction that all will be grue are alike las esmeraldas subsecuentemente examinadas serán
confirmed by evidence statements describing the verdes y también la predicción de que todo será
same observations. But if an emerald "verdul", estarán ambas confirmadas de forma similar,
subsequently examined is grue, it is blue and por declaraciones de evidencia que describirán las
hence not green. Thus although we are well aware mismas observaciones. Pero si una esmeralda
which of the two incompatible predictions is recientemente examinada es de color "verdul",
genuinely confirmed, they are equally well entonces no será azul y por lo tanto tampoco verde.
confirmed according to our present definition. Así aunque estamos adecuadamente al tanto sobre
Moreover, it is clear that if we simply choose an cuál de las dos predicciones incompatibles está
appropriate predicate, then on the basis of these genuinamente confirmada, resulta que están
same observations we shall have equal igualmente bien confirmadas según nuestra definición
confirmation, by our definition, for any prediction presente. Aún más, es claro que si simplemente
whatever about other emeralds-or indeed about escogemos un predicado apropiado, entonces en base
anything else. (10) As in our earlier example, a estas mismas observaciones, nosotros deberemos
only the predictions subsumed under law like contar con una confirmación igual (debido a nuestra
hypothesis are genuinely confirmed; but we have definición) para cualquier predicción no importa qué
no criterion as yet for determining law likeness. se trate de otras esmeraldas --- o incluso sobre
And now we see that without some such criterion, cualquier otra cosa más --- (10). Como en el ejemplo
our definition not merely includes a few anterior, sólo las predicciones subsumidas en la norma
unwanted cilices, but is so completely ineffectual como hipótesis, estarán genuinamente confirmadas;
that it virtually excludes nothing. We are left once Pero no tenemos criterio hasta ahora para una
again with the intolerable result that anything analogía decisiva de norma o ley. Y ahora vemos que
confirms anything. This difficulty cannot be set sin tal criterio, nuestra definición no sólo incluye
aside as an annoying detail to be taken care of in algunos casos indeseados, sino que es tan
due course. It has to be met before our definition absolutamente ineficaz que virtualmente no excluye
will work at all. ninguna cosa. Quedamos otra vez con el intolerable
resultado de que cualquier cosa "confirma lo que sea".
Esta dificultad no puede ser descartada como un
detalle molesto a tener en cuenta en un momento
dado. Tiene que ser encontrada antes de que nuestra
definición se ponga en marcha.

Página 16 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 75

Nevertheless, the difficulty is often slighted No obstarte, la dificultad es a menudo subestimada


because on the surface there seem to be easy porque en su superficie, da la apariencia que de que
ways of dealing with it. Sometimes, for example, existen "maneras fáciles de encargarse de ella". Se
the problem is thought to be much like the piensa que algunas veces, por ejemplo, el problema es
paradox of the ravens. \.Ye are here again, it is muy parecido a la paradoja de los cuervos. Surge de
pointed out, making tacit and illegitimate use of nuevo lo comentado antes, pues se hace un uso tácito
information outside the stated evidence: the e ilegítimo, de información exterior a la evidencia
information, for example, that different samples establecida por ejemplo: La información de que
of one material are usually alike in conductivity, "muestras diferentes de un material se asemejan
and the information that different men in a lecture usualmente en conductividad", o bien la información de
audience are usually not alike in the number of que "hombres diferentes en cierto público no son
their older brothers. But while it is true that such usualmente semejantes en el número de sus
information is being smuggled in, this does not by hermanos mayores". Aunque bien es verdad que tal
itself settle the matter as it settles the matter of the información es internamente arbitraria, esto por sí
ravens. There the point was that when the mismo no resuelve el asunto sobre el cómo encarar la
smuggled information is forth rightly declared, its cuestión de los cuervos. Aquí el punto radica en que
effect upon the confirmation of the hypothesis in cuándo la información arbitraria es dispuesta y
question is immediately and properly registered razonablemente declarada, su efecto en la
by the definition we are using. On the other hand, confirmación de la hipótesis en cuestión es inmediata y
if to our initial evidence we add statements propiamente registrada por la definición que usamos.
concerning the conductivity of pieces of other Por otra parte, si para establecer nuestra prueba inicial
materials or concerning the number of older añadimos declaraciones relacionadas con "la
brothers of members o other lecture audiences, conductividad de piezas de otros materiales", o bien,
this will not in the least affect the confirmation, concernientes al "número de hermanos mayores" que
according to our definition, of the hypothesis tiene la audiencia de otras salas de conferencias,
concerning copper or of that concerning this entonces veremos que esto no afectará finalmente a la
lecture audience. Since our definition is confirmación propuesta en nuestra definición de la
insensitive to the bearing upon hypothesis of hipótesis concerniente al cobre, o sobre la hipótesis
evidence so related to them, even when the relacionada con la audiencia de esta conferencia. Por
evidence is fully declared, the difficulty about lo tanto, ya que nuestra definición es insensible a la
accidental hypothesis cannot be explained away relación o coyuntura de hipótesis de evidencia ligadas
on the ground that such evidence is being entre sí, aun cuando esta evidencia se halla
surreptitiously taken into account. perfectamente declarada, entonces sucederá que la
cuestión relativa a la hipótesis accidental no podrá ser
explicada sobre la base de que tal prueba está siendo
arbitrariamente considerada.

Página 17 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 76

A more promising suggestion is to explain the Una sugerencia más prometedora es explicar el
matter in terms of the effect of this other asunto en términos del efecto de ésta otra evidencia;
evidence not directly upon the hypothesis in no directamente sobre la hipótesis en cuestión sino
question but indirectly through other hypotheses indirectamente a través de otra hipótesis que ya está
that are confirmed, according to our definition, confirmada por tal evidencia, de acuerdo con nuestra
by such evidence. Our information about other definición. Nuestra información acerca de otros
materials does by our definition confirm such materiales por efecto de nuestra definición, confirma
hypotheses as that all pieces of iron conduct hipótesis tales como: "Todo trozo de metal conduce
electricity, that no pieces of rubber do, and so on; electricidad", "Ninguna pieza de hule puede
and these hypotheses, the explanation runs, conducirla" y así continuamente. En estas hipótesis la
impart to the hypothesis that a11 pieces of copper explicación subyacente le da a las hipótesis de que
conduct electricity (and also to the hypothesis "todo trozo de cobre conduce electricidad" (y también
that none do) the character of lawlikeness that is, la hipótesis de que ninguno lo hace) un carácter
amenability to confirmation by direct positive legaliforme --- esto es: se subordina a la confirmación
instances when found. On the other hand, our por instancias positivas directas cuándo nosotros las
information about other lecture audiences encontramos. Por otra parte, nuestra información
disconfirms many hypotheses to the effect that all acerca de la audiencia en otras conferencias, debilita
the men in one audience are third sons, or that muchas hipótesis respecto a que "Todo hombre en la
none are; and this strips any character of audiencia es el hijo menor" (o la hipótesis de que no lo
lawlikeness from the hypothesis that all (or the sea); y esto elimina cualquier carácter de
hypothesis that none) of the men in this audience “legaliformidad” que provenga de la hipótesis de que
are third sons. But clearly if this course is to be “todos los hombres de la audiencia son hijos menores”
followed, the circumstances under which (o la hipótesis de que ninguno lo sea). Pero
hypotheses are thus related to one another will evidentemente si continuamos por este curso,
have to be precisely articulated. entonces las circunstancias bajo la cuales nuestras
hipótesis son relacionadas unas con otras, tendrán
que ser articuladas con más precisión.

Página 18 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN

77

The problem, then, is to define the relevant way El problema entonces, es definir la forma relevante en
in which such hypotheses must be alike. Evidence la cual tales hipótesis deben. asemejarse. Evidencia
for the hypothesis that all iron conducts para la hipótesis de que "todo metal conduce
electricity enhances the lawlikeness of the electricidad" realza la "legaliformidad" de la hipótesis
hypothesis that all zirconium conducts electricity, de que "todo el circonio conduce electricidad", pero no
but does not similarly affect the hypothesis that afecta igual a la hipótesis de que "todos los objetos en
all the objects on my desk conduct electricity. mi escritorio conducen electricidad" ¿Dónde recae la
Wherein lies the difference? The first two diferencia? Las primeras dos hipótesis caen debajo de
hypotheses fall under the broader hypothesis -- otra más amplia --- llamémosla "H" --- la cual dice que
call it "H" --- that every class of things of the "toda clase de cosas del mismo material, es uniforme
same material is uniform in conductivity; the first en conductividad"; la primera y la tercera caen sólo
and third fall only under some such hypothesis as bajo cierta clase de hipótesis --- llamémosle "K" ---
call it "K" that every class of things that are tales como: "toda clase de cosas que sean, o bien del
either all of the same material or all on a desk is mismo material, o bien, que se hallen en un escritorio,
uniform in conductivity. Clearly the important serán uniformes en conductividad". Evidentemente la
difference here is that evidence for a statement diferencia importante aquí es que la evidencia para el
affirming that one of the classes covered by H has enunciado afirmando que alguna de las clases
the property in question increases the credibility incluidas por H, tiene la propiedad en cuestión,
of any statement affirming that another such class incrementa la credibilidad de que cualquier enunciado
has this property; while nothing of the sort holds afirmando que otro de clase semejante tenga esta
true with respect to K. But this is only to say that propiedad; mientras que nada por el estilo puede
H is lawlike and K is not. We are faced anew conservar su validez con relación a K. Pero esto es
with the very problem we are trying to solve: the sólo para decir que H es "legaliforme" y “K” no lo es.
problem of distinguishing between lawlike and Somos confrontados nuevamente con el mismo
accidental hypotheses. problema que tratamos de resolver: El problema de
distinguir lo legaliforme de la hipótesis accidental.

Página 19 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
78
NELSON GOODMAN

The most popular way of attacking the problem La forma más popular de atacar el problema se origina
takes its cue from the fact that accidental del hecho de que una hipótesis accidental parece
hypotheses seem typically to involve some spatial envolver típicamente algunas restricciones espaciales
or temporal restriction, or reference to some o temporales, o bien, la referencia a algún individuo
particular individual. They seem to concern the particular. por ejemplo, esto parece tener que ver con
people in some particular room, or the objects on las personas en un cuarto en particular o los objetos
some particular person's desk; while lawlike en el escritorio de una persona en particular, mientras
hypotheses characteristically concern all ravens que una hipótesis "legaliforme" le concierne
or all pieces of copper whatsoever. Complete característicamente "a todos los cuervos" o "a todos
generality is thus very often supposed to be a los objetos de cobre, cualquiera que estos sean". La
sufficient condition of lawlikeness; but to define completa generalidad suele suponerse como condición
this complete generality is by no means easy. necesaria de la "legaliformidad"; pero definir esta
Merely to require that the hypothesis contain no "completa generalidad" no es de ninguna manera fácil.
term naming, describing, or indicating a particular Simplemente exigir que la hipótesis no contenga
thing or location will obviously not be enough. ningún término nombrando, describiendo o indicando
The troublesome hypothesis that all emeralds are una cosa o locación en particular, obviamente no será
grue contains no such term; and where such a suficiente. La hipótesis problemática de que todas las
term does occur, as in hypotheses about men in esmeraldas son "verdul". no contiene tal término y
this room, it can be suppressed in favor of some donde este aparece y donde tal término surge, como
predicate (short or long, new or old) that contains en la hipótesis acerca de los hombres en esta sala,
no such term but applies only to exactly the same este puede suprimirse a favor de algún predicado que
things. One might think, then, of excluding not no contiene tal término pero se aplica sólo y
only hypotheses that actually contain terms for exactamente a las mismas cosas (corto o largo, nuevo
specific individuals but also all hypotheses that o viejo). Uno podría pensar entonces, en excluir no
are equivalent to others that do contain such sólo hipótesis que realmente contienen términos para
terms. But, as we have just seen, to exclude only personas físicas específicas, sino también en excluir
hypotheses of which all equivalents contain such toda hipótesis que sea equivalente para términos que
terms is to exclude nothing. On the other hand, to contengan tales condiciones. Pero, como
exclude all hypotheses that have some equivalent precisamente se ha visto, excluir una hipótesis en la
containing such a term is to exclude everything; cual todos los equivalentes contienen tales
for even the hypothesis condiciones, es lo mismo que no excluir ninguna cosa.
Por el otro lado, excluir toda hipótesis que encierre un
equivalente conteniendo tal término es excluirlo todo;
Comparando las hipótesis tenemos:

All grass is green Todo pasto es verde


has as an equivalent tiene como equivalente
All grass in London or elsewhere is green. todo pasto en Londres o cualquier otro lugar es verde

Página 20 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN

NELSON GOODMAN 79

The next step, therefore, has been to consider El siguiente paso, entonces, es reflexionar
ruling out predicates of certain kinds. A descartando predicados de ciertos tipos. Una hipótesis
syntactically universal hypothesis is lawlike, the sintácticamente universal es "legaliforme" (dice
proposal runs, if its predicates are 'purely nuestra propuesta) si sus predicados son "puramente
qualitative' or 'non - positional'. (11) This will cualitativos" ' o “no-posicionales”. (11) Entonces esto,
obviously accomplish nothing if a purely obviamente, no logrará ninguna cosa si un predicado
qualitative predicate is then conceived either as puramente cualitativo es concebido ya sea como uno
one that is equivalent to some expression free of que es equivalente a alguna expresión libre de
terms for specific individuals, or as one that is condiciones para individuos específicos, o como uno
equivalent to no expression that contains such a que no es equivalente a ninguna expresión que
term; for this only raises again the difficulties just contenga tal término; pues esto sólo pone a la vista
pointed out. The claim appears to be rather that at otra vez las dificultades recién apuntadas. La
least in the case of a simple enough predicate we pretensión parece ser más bien, que al menos en el
can readily determine by direct inspection of its caso de un predicado simple y suficiente, nosotros
meaning whether or not it is purely qualitative. fácilmente podemos determinar a través de la
But even aside from obscurities in the notion of inspección directa de su significado, si es o no
'the meaning' of a predicate, this claim seems to puramente cualitativo. Pero aun además de
me wrong. I simply do not know how to tell oscuridades en la noción del significado de un
whether a predicate is qualitative or positional, predicado, esta demanda me parece equivocada.
except perhaps by completely begging the Simplemente no sé cómo decir si un predicado es
question at issue and asking whether the predicate cualitativo o posicional, excepto quizá a través de
is 'well-behaved'-that is, whether simple evadir completamente el asunto sobre la pregunta y la
syntactically universal hypothesis applying it are respuesta de si el predicado es "adecuadamente
lawlike. comportado" - esto es, si las hipótesis universales
sintácticamente simples aplicadas son "legaliformes".

This statement will not go unprotested. Esta afirmación amerita un examen. “Supóngase” que
"Consider", it will be argued, "the predicates se argumentaran los predicados "azul" y 'verde' y el
'blue' and 'green' and the predicate “grue” predicado 'verdul' introducido antes; además piense
introduced earlier, and also the predicate 'bleen' también en el predicado 'bleen' que se aplica a las
that applies to emeralds examined before time t esmeraldas examinadas antes del tiempo "t" siendo
just in case they are blue and to other emeralds entonces azules y las otras verdes. Con seguridad
just in case they are green. Surely it is clear", the está claro, nos dice el argumento, que las primeras
argument runs, "that the first two are purely dos son puramente cualitativas y las siguientes dos, no
qualitative and the second two are not; for the lo son; Pues el significado de cada una de las últimas
meaning of each of the latter two plainly involves dos explícitamente implica referencias a posiciones
reference to'a specific temporal position." To this temporales específicas A esto yo contesto que
I reply that indeed I do recognize the first two as ciertamente reconozco a las primeras dos como
well-behaved predicates admissible in lawlike "predicados apropiadamente comportados" admisibles
hypotheses, and the second two as ill-behaved en hipótesis legaliformes, y los dos posteriores como
predicates. But the argument that the former but "predicados inapropiadamente comportados". Pero el
not the latter are purely qualitative seems to me argumento de que las primeras pero no las posteriores
quite unsound. True enough, if we start with sean puramente cualitativas, me parece muy
"blue" and "green", then "grue" and "bleen" will infundado. Cierto es que, si comenzamos con "azul" y

Página 21 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
80
NELSON GOODMAN

be explained in terms of "blue" and "green" and a "verde", pues entonces "verdul" y "bleen" estarán
temporal term. But equally truly, if we start with explicados en términos de "azul" y "verde" y de un
"grue" and "bleen", then "blue" and "green" will término temporal. Pero si todo fuera verdadero, y
be explained in terms of digrue" and "bleen" and comenzáramos con "verdul" y "bleen" para que luego
a temporal term; 66 green", for example, applies "el azul" y "el verde" estén explicados en términos de
to emeralds examined before time t just in case "verdul" y "bleen" junto a un término temporal;
they are grue, and to other emeralds just in case "verdes", por ejemplo, se aplicaría a las esmeraldas
they are bleen. Thus qualitativeness is an entirely examinadas antes del tiempo "t" y justo en ese caso
relative matter and does not by itself establish any sería verdul, y las otras esmeraldas justo en este caso,
dichotomy of predicates. This relativity seems to serían "bleen". Así lo cualitativo es un asunto
be completely overlooked by those who contend enteramente relativo y por sí mismo no establece
that the qualitative character of a predicate is a alguna dicotomía de predicados. Esta relatividad
criterion for its good behavior. parece estar completamente descuidada por aquéllos
que consideran que el carácter cualitativo de un
predicado es un criterio para un "comportamiento
apropiado".

Of course, one may ask why we need worry about Por supuesto alguien podría preguntar por qué
such unfamiliar predicates as "grue" or about necesitamos preocuparnos acerca de tales predicados
accidental hypotheses in general, since we are tan poco familiares como "verdul" o de hipótesis
unlikely to use them in making predictions. If our accidentales en general, ya que es poco probable que
definition works for such hypotheses as are los usemos al realizar alguna predicción. Si nuestra
normally employed, isn't that all we need? In a definición para tales hipótesis son normalmente
sense, yes; but only in the sense that we need no empleadas ¿No son todo lo que necesitamos? En
definition, no theory of induction, and no cierto sentido, sí, pero sólo en el sentido de que no
philosophy of knowledge at all. We get along necesitemos en absoluto de ninguna definición,
well enough without them in daily life and in ninguna teoría de la inducción y ninguna filosofía del
scientific research. But if we seek a theory at all, Conocimiento. Conseguimos sobrevivir lo
we cannot excuse gross anomalies resulting from suficientemente bien sin ellos en la vida diaria y en la
a proposed theory by pleading that we can avoid investigación científica. Pero si buscamos una teoría
them in practice. The odd cases we have been del todo, entonces no podemos excusar anomalías
considering are clinically pure cases that, though vulgares que resultan de una teoría propuesta,
seldom encountered in practice, nevertheless pretextando que las podemos evitar en la práctica. Los
display to best advantage the symptoms of a casos extraños que hemos estado considerando aquí
widespread and destructive malady. son casos "clínicamente puros" que, aunque rara vez
encontrados en la práctica, tienen la ventaja de
evidenciar mejor los síntomas de un mal extendido y
destructivo.

Página 22 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
81
NELSON GOODMAN

We have so far neither any answer nor any No tenemos hasta ahora alguna respuesta ni alguna
promising clue to an answer to the question what pista prometedora para una respuesta a la pregunta
distinguishes lawlike or confirmable hypothesis sobre lo que distingue una hipótesis confirmable ó
from accidental or no - confirmable ones; and "legaliforme", de otra accidental o "no-confirmable"; y
what may at first have seemed a minor technical lo que al principio se miró como una dificultad técnica
difficulty has taken on the stature of a major menor, ha cobrado la estatura de un obstáculo
obstacle to the development of a satisfactory principal para el desarrollo de una teoría satisfactoria
theory of confirmation. It is this problem that I de la confirmación. Este es el problema que yo llamo:
call the new riddle of induction. "El nuevo problema de la Inducción".

5. The Pervasive Problem of Projection 5. El problema generalizado de la Proyección.

At the beginning of this lecture, I expressed the Al principio de esta lectura, yo expresé la opinión de
opinion that the problem of induction is still que el problema de la inducción todavía permanece sin
unsolved, but that the difficulties that face us resolverse, además, que las dificultades que
today are not the old ones; and I have tried to enfrentamos hoy no son las mismas que las antiguas;
outline the changes that have taken place. The y que trataría de subrayar los cambios que habían
problem of justifying induction has been acontecido. El problema de "justificar a la inducción"
displaced by the problem of defining ha sido desplazado por el problema de "Definir la
confirmation, and our work upon this has left us confirmación", y nuestro trabajo sobre el tema, nos ha
with the residual problem of distinguishing dejado con el problema residual de distinguir entre
between confirmable and non-confirmable hipótesis confirmable y no-confirmables. Uno podría
hypotheses. One might say roughly that the first decir con razón que la primera pregunta fue: ¿Por qué
question was "Why does a positive instance of a la instancia positiva de una hipótesis puede llegar a
hypothesis give any grounds for predicting further establecer algún fundamento para la predicción de
instances?"; that the newer question was "What is otras instancias?; y que la nueva cuestión era: "¿Qué
a positive instance of a hypothesis?"; and that the es la "instancia positiva" de una hipótesis?"; y además
crucial remaining question is "What hypotheses que la pregunta crucial es: ¿Cuáles hipótesis son
are confirmed by their positive instances?" confirmadas a través de sus "instancias positivas"?

The vast amount of effort expended on the Hay que decir que el vasto conjunto de esfuerzos
problem of induction in modern times has thus invertidos en el problema de la Inducción en los
altered our afflictions but hardly relieved them. tiempos modernos, han alterado nuestras aflicciones,
The original difficulty about induction arose from pero difícilmente han logrado removerlas. La dificultad
the recognition that anything may follow upon original acerca de la inducción surge del
anything. Then, in attempting to define reconocimiento de que cualquier cosa puede ser
confirmation in terms of the converse of the consecuencia de cualquier otra cosa. Entonces, en un
consequence relation, we found ourselves with intento de definir la confirmación en términos de "una
the distressingly similar difficulty that our inversa" de la relación de consecuencia, nos
definition would make any statement confirm any encontramos con la dificultad angustiosamente similar,
other. And now, after modifying our definition de que nuestra definición haría que cualquier
drastically, we still get the old devastating result declaración confirmase cualquier otra. Y todavía
that any statement will confirm any statement. después de modificar nuestra definición drásticamente,
Until we find a way of exercising some control aún obtenemos el antiguo y devastador resultado de
over the hypotheses to be admitted, our definition que cualquier declaración confirmará cualquier otra
makes no distinction whatsoever between valid declaración. Entonces, hasta que encontremos una

Página 23 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN 82
NELSON GOODMAN

and invalid inductive inferences. forma de ejercer algún control sobre las hipótesis a ser
admitidas, nuestra definición no hará ninguna
distinción entre las inferencias inductivas válidas e
inválidas.

The real inadequacy of Hume's account lay not in La verdadera inconveniencia del argumento de Hume
his descriptive approach but in the imprecision of descansa, no en su enfoque descriptivo, sino en la
his description. Regularities in experience, imprecisión de su descripción. Las regularidades en la
according to him, give rise to habits of experiencia, de a cuerdo con él, da origen a ciertos
expectation; and thus it is predictions conforming hábitos de expectativa; y es así como las predicciones
to past regularities that are normal or valid. But hechas de acuerdo a regularidades pasadas son
Hume overlooks the fact that some regularities do normales o válidas. Pero Hume pasa por alto el hecho
and some do not establish such habits; that de que algunas regularidades establecen tales hábitos,
predictions based on some regularities are valid y otras no. Cada palabra que usted me ha oído decir
while predictions based on other regularities are ha ocurrido antes de la frase final de esta conferencia;
not. Every word you have heard me say has Pero no se genera, espero, alguna expectativa
occurred prior to the final sentence of this lecture; especial por el hecho de que las palabras que usted
but that does not, I hope, create any expectation escucha, estén antes de la frase final. Otra vez,
that every word you will hear me say will be prior considere nuestro caso de las esmeraldas. Todas
to that sentence. Again, consider our case of aquellas examinadas antes del tiempo "t" son verdes,
emeralds. All those examined before time t are y nos permite esperar y confirmar la predicción de que
green; and this leads us to expect, and confirms la siguiente sea verde. Pero sucede que también,
the prediction, that the next one will be green. But todas las examinadas son "verdules" y esto no nos
also, all those examined are grue; and this does deja esperar, ni tampoco confirmar la predicción de
nor lead us ro expect, and does not confirm the que la siguiente gema será "verdul". La regularidad en
prediction, that the next one will be grue. el verdor confirma la predicción en ciertos casos, la
Regularity in greenness confirms the prediction of regularidad de la "verdules" no lo hace. Decir que las
further cases; regularity in grueness does not. To predicciones válidas son aquellas basadas en
say that valid predictions are those based on past regularidades pasadas, sin ser capaz de explicar
regularities, without being able to say which cuáles son esas regularidades, es por lo tanto, algo
regularities, is thus quite pointless. Regularities bastante superfluo. Las regularidades están donde
are where you find them, and you can find them usted las encuentre, y usted las puede encontrar
anywhere. As we have seen, Hume's failure to donde sea. Como lo hemos visto, la falla de Hume
recognize and deal with this problem has been para reconocer y tratar con este problema, ha sido
shared even by his most recent successors. compartida incluso por sus sucesores más recientes.

As a result, what we have in current confirmation Como consecuencia, lo que tenemos en la actual
theory is a definition that is adequate for certain teoría de la confirmación, es una definición que es
cases that so far can be described only as those adecuada para ciertos casos que hasta ahora sólo
for which it is adequate. The theory works where pueden ser descritos "como esos para los cuales es
it works.'^ hypothesis is confirmed by statements adecuada". Entonces la teoría surte efecto
related to it in the prescribed way provided it is so dondequiera que sea. La hipótesis es confirmada por
confirmed. This is a good deal like having a declaraciones relacionadas a ella en la forma prescrita,
theory that tells us that the area of a plane figure y así, ella es confirmada. Esto es tan conveniente
is one half the base times the altitude, without como tener una teoría que nos dice que el área de la
telling us for what figures this holds. We must figura plana de un avión es "la mitad de la base por la

Página 24 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN
83
NELSON GOODMAN

somehow find a way of distinguishing lawlike altitud", sin especificarnos cuáles son las figuras que la
hypothesis, to which our definition of construyen. De alguna manera debemos encontrar una
confirmation applies, from accidental hypothesis, forma de distinguir la hipótesis "legaliformes", para que
to which it does not. nuestra definición de confirmación separe las hipótesis
accidentales de las que no lo son.

Today I have been speaking solely of the problem Hoy estuve hablando con calma del problema de la
of induction, but what has been said applies inducción, pero lo que he estado diciendo aplica por
equally to the more general problem of igual al problema más general de la proyección. Como
projection. As pointed out earlier, the problem of lo señalé antes, el problema de la proyección que va
prediction from past to future cases is but a de los casos pasados a los futuros no es sino una
narrower version of the problem of projecting versión especial del problema de proyectar cualquier
from any set of cases to others. We saw that a escenario de casos a otros distintos. Vimos que un
whole cluster of troublesome problems grupo entero de problemas difíciles, concernientes a
concerning dispositions and possibility can be las disposiciones y a la posibilidad, pueden reducirse
reduced to this problem of projection. That is why al problema de la proyección. Este el por qué
the new riddle of induction, which is more consideramos que el nuevo enigma de la inducción (el
broadly the problem of distinguishing between cual es más ampliamente el problema de distinguir
projectible and non-projectible hypotheses, is as entre la hipótesis proyectable y la no proyectadle), es
important as it is exasperating. tan importante como también exasperante.

Our failures teach us, I think, that lawlike or Los errores nos enseñan, pienso, que la hipótesis
projectible hypotheses cannot be distinguished on legaliforme o proyectable, no puede ser distinguida
any merely syntactical grounds or even on the sobre la sola base de argumentos sintácticos, o aun
ground that these hypotheses are somehow purely sobre la base de que éstas hipótesis son, en cierta
general in meaning. Our only hope lies in re- forma, de una carácter puro y general en su
examining the problem once more and looking significado. Nuestra única esperanza consiste en re-
for some new approach. This will be my course in examinar el problema otra vez y buscar algún
the final lecture. acercamiento nuevo. Ésta será mi dirección en la
conferencia final.

Página 25 de 25 Hecho, Ficción y Predicción - Capítulo III


EL NUEVO ENIGMA DE LA INDUCCIÓN NELSON GOODMAN

Hecho, Ficción y Predicción - Capítulo III

NOTAS CAPÍTULO III


1

1. Although this remark is merely an aside, perhaps 1. Aunque este comentario es meramente una digresión,
I should explain for the sake of some unusually quizá debería explicar por el bien de nuestros lectores más
sheltered reader that the notion of a necessary conservadores que la noción de una conexión necesaria
connection of ideas, or of an absolutely analytic de ideas, o de una declaración absolutamente analítica,
statement, is no longer sacrosanct. Some, like ya no es algo "sacrosanto". Algunos como Quine y White,
Quine and White, have forth rightly attacked the últimamente han atacado la noción. Otros como yo,
notion; others, like myself, have simply discarded it; simplemente estamos en desacuerdo con ella .y todavía
and still others have begun to feel acutely otros han comenzado a sentirse extremadamente
uncomfortable about it. incómodos respecto al tema.
2. A hasty reader might suppose that my insistence 2. Un lector apresurado podría suponer que mi insistencia
here upon identifying the problem of justification aquí al identificar el problema de justificación con un
with a problem of description is out of keeping with problema de descripción está relacionada con mi
my parenthetical insistence in the preceding lecture insistencia parentética en la disertación precedente
that the goal of philosophy is something quite respecto a que la meta de la filosofía es algo muy diferente
different from the mere description of ordinary or de la mera descripción de un procedimiento ordinario o
scientific procedure. Let me repeat that the point científico. Permítaseme repetir que el punto discutido fue
urged there was that the organization of the que la organización de un recuento aclaratorio no necesita
explanatory account need not reflect the manner or reflejar la manera o el orden en el cual los predicados son
order in which predicates are adopted in practice. It adoptados en la práctica. Seguramente debe describir la
surely must describe practice, however, in the sense práctica, sin embargo, en el sentido de que las
that the extensions of predicates as explicated must extensiones de predicados como lo hemos explicado
conform in certain ways to the extensions of the same deben regularse en cierta forma a las extensiones de los
predicates as applied in practice. Hume's account is a mismos predicados como son aplicados en la práctica. La
description in just this sense. For it is an attempt to relación de Hume es una descripción justamente este
set forth the circumstances under which those sentido. Pues es un intento de exponer las condiciones
inductive judgments are made that are normally bajo las cuales esos juicios inductivos están hechos que
accepted as valid; and to do that is to state necessary son normalmente aceptadas como válidas; Y hacer eso es
and sufficient conditions for, and thus to define, valid declarar condiciones necesarias y suficientes a favor de ( y
induction. What I am maintaining above is that the también en pro de definir), una inducción válida. Lo que
problem of justifying induction is not something over sostengo arriba está que el problema sobre la justificación
and above the problem of describing or defining de la inducción no es algo que se encuentre más allá del
valid induction. problema de la descripción o la definición de una inducción
válida.

3 For a fuller discussion of definition in general see 3 Para un debate más completo sobre la definición en
Chapter I of The Structure of Appearance. general véase el Capítulo sobre La Estructura de la
Apariencia.
4. The basic article is 'A Purely Syntactical 4. El artículo básico es '"una definición meramente
Definition of Confirmation', cited in Note 1.10. A sintáctica de la Confirmación", citada en la nota 1.10. Una
much less technical account is given in 'Studies in the explicación mucho menos especializada es dada en
Logic of Confirmation', Mind, n.s., vol.. 54 (1945) "Estudios sobre la Lógica de la Confirmación", Mind, n.s.,
pp. 1-26 and 97-121. Later work by Hempel and vol.. 54 (1945) pp.1-26 y 97-121. El trabajo posterior de
others on defining degree of confirmation does not Hempel y otros autores sobre grado de definición de la
concern us here. confirmación no nos concierne aquí.

Traducción: Madrigal Bailón Joel Tucídides / UAM-iztapalapa / IN CALLI IXCAHUICOPA


EL NUEVO ENIGMA DE LA INDUCCIÓN NELSON GOODMAN

Hecho, Ficción y Predicción - Capítulo III

5. 1 am not here asserting that this is an 5. No estoy afirmando aquí que este sea un requisito
indispensable requirement upon a definition of indispensable para la definición de confirmación. Debido a
confirmation. Since our commonsense assumptions que nuestras suposiciones de sentido común, tomadas en
taken in combination quickly lead us to absurd combinación, rápidamente nos conducen hacia
conclusions, some of these assumptions have to be conclusiones absurdas, algunas de estas suposiciones
dropped; and different theorists may make different tienen que ser descartadas; Y cada teórico debe tomar
decisions about which to drop and which to preserve. decisiones diferentes sobre cuales descartar y cuáles a
Hempel gives up the converse consequence conservar. Hempel descarta la condición "inversa" de la
condition, while Carnap (Logical Foundations of consecuencia, mientras que Carnap descarta ambas: la
Probability, Chicago and London, 1950, pp. 474-6) condición de consecuencia y la condición opuesta de
drops both the consequence condition and the consecuencia (Las Bases Lógicas de la Probabilidad,
converse consequence condition. Such differences of Chicago y Londres, 1950, pp. 474-6). Tales diferencias de
detail between different treatments of confirmation detalle entre los tratamientos diferentes de confirmación no
do not affect the central points I am making in this afectan las proposiciones centrales que hago en esta
lecture. conferencia.
6. Any hypothesis is 'supported' by its own positive 6. Cualquier hipótesis se sostiene a través de instancias
instances; but support-or better, direct factual positivas muy propias, pero el soporte --- mejor dicho --- el
support-is only one factor in confirmation. This soporte factual directo es el único factor bajo confirmación.
factor has been separately studied by John G. Este factor ha sido aisladamente estudiado por John G .
Kemeny and Paul Oppenheim in 'Degree of Factual Kemeny y Paul Oppenheim en "El Grado de soporte
Support', Philosophy of Science, vol. 19 (1952)~pp . factual", Filosofía de la Ciencia vol. 19 pp. (1952) 307-24.
307-24. AS will appear presently, my concern in como indica este trabajo, mi preocupación en estas
these lectures is primarily with certain other conferencias es ante todo debida a ciertos factores
important factors in confirmation, some of them quite especiales e importantes en la confirmación, algunos de
generally neglected. ellos generalmente descuidados.
7. An able and thorough exposition of this paragraph 7. Una posible y completa exposición de este parágrafo es
is given by Israel Scheffler in his Anatomy of dado por Israel Scheffler es su "Anatomía de la
Inquiry, New York, 1963, pp. 286-91. Investigación" New York, 1963, pp. 286-91.
8. The status of the conjunction condition is much 8. El estatus de la condición de la conjunción es muy
like that of the consequence condition-see Note parecida a la de una condición de consecuencia --- véase
111.5. Although Carnap drops the cbnjunction Nota III.5 . Aunque Carnap descarta también la condición
condition also (p. 394), he adopts for different de conjunción (p. 394), él adopta por distintos motivos, la
reasons the requirement we find needed above: that exigencia que hemos encontrado como necesaria arriba:
the total available evidence must always be taken que la evidencia total disponible siempre tiene que
into account (pp.21 1-13). tomarse en cuenta (pp. 211-213)
9. Although the example used is different, the 9. Aunque el ejemplo usado es diferente, la discusión a
argument to follow is substantially the same as that seguir está sustancialmente la misma que establecí en mi
set forth in my note 'A Query on Confirmation', cited ponencia "Una consulta sobre la confirmación" citada en la
in Note 1.16. nota 1.16.

Traducción: Madrigal Bailón Joel Tucídides / UAM-iztapalapa / IN CALLI IXCAHUICOPA


EL NUEVO ENIGMA DE LA INDUCCIÓN NELSON GOODMAN

Hecho, Ficción y Predicción - Capítulo III

3
10. For instance, we shall have equal confirmation, 10.- Por ejemplo, debemos tener una confirmación similar
by our present definition, for the prediction that roses valiéndonos de nuestra definición presente, para la
subsequently examined will be blue. Let "emerose" predicción de que: "Todas las rosas posteriormente
apply just to emeralds examined before time t, and to examinadas serán azules". Apliquemos ahora la palabra
roses examined later. Then all emeroses so far "emerose" a las esmeraldas examinadas antes del tiempo
examined are grue, and this confirms the hypothesis "t" y a las rosas examinadas después de él. Entonces
that all emeroses are grue and hence the prediction todos los "emeroses" hasta ahora examinados son "grue",
that roses subsequently esamined will be blue. The y esto confirma la hipótesis que todos los emeroses son
problem raised by such antecedents has been little "grue" y por lo tanto la predicción que las rosas
noticed, but is no easier to meet than that raised by posteriormente examinadas serán azules. El problema
similarly perverse consequences. See further IV,4 surgido de tales antecedentes ha sido poco advertido, pero
below. no es más fácil de afrontar que el surgido por
consecuencias adversas de caracter semejante. Vea
además el punto IV.4 más adelante.
11. Carnap took this course in his paper 'On the 11. Carnap tomó este curso en su escrito "Sobre la
Application of Inductive Logic', Philosophy and Aplicación de la Lógica Inductiva", Investigación Filosófica
Phenomenological Research, VOI. 8 ( 1947), pp. y Fenomenológica. Vol. 8 (1947), pp. 133-47, la cual es en
133-47, which is in part a reply to my 'A Query on parte una réplica a mi trabajo "Una Pregunta sobre la
Confirmation', cited in Note 1.16. The discussion was Confirmación" citada en la nota 1.16 La discusión fué
continued in my note 'On Infirmities of Confirmation continuada en mi ensayo: "Sobre las Debilidades de la
Theory', Philosophy and Phenomenological Teoría de la Confirmación", Investigación Filosófica y
Research, vol. 8 (1947) p p. 149-5 1 ; and in Carnap's Fenomenológica, vol. 8 (1947), pp. 149-51 y en el artículo
'Reply to Nelson Goodman', same journal, same de Carnap "Réplica a Nelson Goodman", misma revista,
volume, pp. 461-2. mismo volumen, p. 461-62.

Tema: Goodman’s New Riddle of Induction. Traducido por Joel Tucídides Madrigal Bailón, Licenciado en Filosofía,
México Distrito Federal. Universidad Autónoma Metropolitana Unidad Iztapalapa (tucidides2000@yahoo.com.mx).
Para educar y sin fines de lucro. Se permite la reproducción total o parcial citando al autor original.

Dedicado con admiración a los estudiantes que tratan de entender y se esfuerzan por un futuro mejor. Invito a los
amigos a que colaboren generosamente con sus escritos traducidos en bien de las generaciones futuras.

Artículo inglés de dominio público hallado en red el 15 de mayo del 2011 del vínculo: google [PDF] THE NEW RIDDLE
OF INDUCTION . Agradecemos a este valioso colaborador anónimo su amable escaneo. Esta versión castellana
puede descargarse de la página: http://es.scribd.com/ ubicando “Nelson Goodman el nuevo enigma de la
inducción” y en www.losriosdeheraclito.blogspot.com .

Referencia bibliográfica: "The New Riddle of Induction." Chapter III of Nelson Goodman (1983): Fact, Fiction, and
Forecast. Cambridge, MA: Harvard University Press, 59-83.

Para Angélica, porque su cariño e inteligencia son mi fuerza en


la adversidad.

Para Itzel Ruíz Madrigal, porque su sonrisa es la prueba del buen futuro que
todavía está por venir.

http://web.me.com/delicious_monsters/N.Ritcey/Philosophy_111a_files/The%20New%20Riddle%20of%20Induction.p
df

Traducción: Madrigal Bailón Joel Tucídides / UAM-iztapalapa / IN CALLI IXCAHUICOPA


THE PASSING OF THE POSSIBLE

feh to non-manifest cases is obviously not very different


from the problem of going from known to unknown or
from past to future cases. The problem of dispositions
looks suspiciously like one of the philosopher's oldest
friends and enemies: the problem of induction. Indeed,
the two are but different aspects of the general problem
of proceeding from a given set of cases to a wider set. The
critical questions throughout are the same: when, how, T H E N E W RIDDLE OF INDUCTION
why is such a transition or expansion legitimate? In the
next lecture, then, we must see how matters stand at I. T h e Old Problem of Indz~ction
present with the familiar problem of induction.
At the close of the preceding lecture, I said that today I
Thus passes the possible. It passes, indeed, only into
another and exceedingly difficult problem. But that prob- should examine how matters stand with respect to the
lem has been troubling our sleep for a long time on its own problem of induction. In a word, I think they stand ill.
But the real difficulties that confront us today are not the
account. There is perhaps some solace in the thought that
traditional ones. What is commonly thought of as the
at least the ghost of the possible will no longer be thump-
ing in the attic. Problem of Induction has been solved, or dissolved; and
we face new problems that are not as yet very widely
understood. T o approach them, I shall have to run as
quickly as possible over some very familiar ground.
The problem of the validity of judgments about future
or unknown cases arises, as Hume pointed out, because
such judgments are neither reports of experience nor logi-
cal consequences of it. Predictions, of course, pertain to
what has not yet been observed. And they cannot be
logically inferred from what has been observed; for what
has happened imposes no logical restrictions on what will
happen. Although Hume's dictum that there are no neces-
sary connections of matters of fact has been challenged
at times, it has withstood all attacks. Indeed, I should be
inclined not merely to agree that there are no necessary
connections of matters of fact, but to ask whether there
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION
\
are any necessary connections at all1-but that is another question of our license for making them. T o trace origins,
story. runs the old complaint, is not to establish validity: the
Hume's answer to the question how predictions are real question is not why a prediction is in fact made but
related to past experience is refreshingly non-cosmic. how it can be justified. Since this seems to point to the
When an event of one kind frequently follows upon an awkward conclusion that the greatest of modern philoso-
event of another kind in experience, a habit is formed that phers completely missed the point of his own problem, the
leads the mind, when confronted with a new event of the idea has developed that he did not really take his solution
first kind, to pass to the idea of an event of the second very seriously, but regarded the main problem as unsolved
kind. The idea of necessary connection arises from the and perhaps as insoluble. Thus we come to speak of
felt impulse of the mind in making this transition. 'Hume's problem' as though he propounded it as a ques-
Now if we strip this account of all extraneous features, tion without answer.
the central point is that to the question "Why one pre- All this seems to me quite wrong. I think Hume grasped
diction rather than another?", Hume answers that the the central question and considered his answer to be
elect prediction is one that accords with a past regularity, passably effective. And I think his answer is reasonable
because this regularity has established a habit. Thus and relevant, even if it is not entirely satisfactory. I shall
among alternative statements about a future moment, one explain presently. At the moment, I merely want to record
statement is distinguished by its consonance with hibit a protest against the prevalent notion that the problem of
and thus with regularities observed in the past. Prediction justifying induction, when it is so sharply dissociated
according to any other alternative is errant. from the problem of describing how induction takes place,
H o w satisfactory is this answer? The heaviest criticism can fairly be called Hume's problem.
has taken the righteous position that Hume's account at I suppose that the problem of justifying induction has
best pertains only to the source of predictions, not their called forth as much fruitless discussion as has any half-
legitimacy; that he sets forth the circumstances under way respectable problem of modern philosophy. The
which we make given predictions-and in this sense ex- typical writer begins by insisting that some way of justi-
plains why we make them-but leaves untouched the fying predictions must be found; proceeds to argue that
for this purpose we need some resounding universal law
1 Although this remark is merely an aside, perhaps I should ex- of the Uniformity of Nature, and then inquires how this
plain for the sake of some unusually sheltered reader that the no- universal principle itself can be justified. At this point, if
tion of a necessary connection of ideas, or of an absolutely an-
alytic statement, is no longer sacrosanct. Some, like Quine and
he is tired, he concludes that the principle must be ac-
White, have forthrightly attacked the notion; others, like myself, cepted as an indispensable assumption; or if he is energetic
have simply discarded it; and still others have begun to feel acutely and ingenious, he goes on to devise some subtle justifica-
uncomfortable about it. tion for it. Such an invention, however, seldom satisfies
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION
7

anyone else; and the easier course of accepting an unsub- standing of our problem can be gained by looking for a
stantiated and even dubious assumption much more sweep- moment at what is involved in justifying non-inductive
ing than any actual predictions we make seems an odd and inferences. How do we justify a deduction? Plainly, by
expensive way of justifying them. showing that it conforms to the general rules of deductive
inference. An argument that so conforms is justified or
2. Dissolutio~zof the Old Problem valid, even if its conclusion happens to be false. An argu-
ment that violates a rule is fallacious even if its conclusion
Understandably, then, more critical thinkers have sus- happens to be true. T o justify a deductive conclusion
pected that there might be something awry with the prob- therefore requires no knowledge of the facts it pertains to.
lem we are trying to solve. Come to think of it, what Moreover, when a deductive argument has been shown to
precisely would constitute the justification we seek? If conform to the rules of logical inference, we usually con-
the problem is to explain how we lmow that certain pre- sider it justified without going on to ask what justifies the
dictions will turn out to be correct, the sufficient answer rules. Analogously, the basic task in justifying an inductive
is that we don't h o w any such thing. If the problem is to inference is to show that it conforms to the general rules
fi77d some way of distinguishing antecedently between of induction. Once we have recognized this, we have gone
true and false predictions, we are asking for prevision a long way towards clarifying our problem.
rather than for philosophical explanation. Nor does it help Yet, of course, the rules themselves must eventually be
matters much to say that we arc merely trying to show justified. The validity of a deduction depends not upon
that or why certilin predictions are probable. Often it is conformity to any purely arbitrary rules we may contrive,
said that while we cannot tell in advance whether a pre- but upon conformity to valid rules. When we speak of
diction concerning a given throw of it die is true, we can the rules of inference we mean the valid rules-or better,
decide whether the prediction is a probable one. But if this some valid rules, since there may be alternative sets of
means determining how the prediction is related to actual equally valid rules. But how is the validity of rules to be
frequency distributions of future throws of the die, surely determined? Here again we encounter philosophers who
there is no way of knowing or proving this in advance. insist that these rules follow from some self-evident axiom,
O n the other hand, if the judgment that the prediction is and others who try to show that the rules are grounded
probable has nothing to do with subsequent occurrences, in the very nature of the human mind. I think the answer
then the question remains in what sense a probable pre- lies much nearer the surface. Principles of deductive in-
diction is any better justified than an inlprobable one. ference are justified by their conformity with accepted
Now obviously the genuine problem cannot be one of deductive practice. Their validity depends upon accord-
attaining unattainable knowledge or of accounting for ance with the particular deductive inferences we actually
knowledge that we do not in fact have. A better under- make and sanction. If a rule yields inacceptable inferences,
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION

we drop it as invalid. Justification of general rules thus accepted inductive judgments are made, he was in fact
derives from judgments rejecting or accepting particular dealing with the question of inductive ~ a l i d i t y .The
~
deductive inferences. validity of a prediction consisted for him in its arising
This looks flagrantly circular. I have said that deductive from habit, and thus in its exemplifying some past regu-
inferences are justified by their conformity to valid gen- larity. His answer was incomplete and perhaps not en-
eral rules, and that general rules are justified by their con- tirely correct; but it was not beside the point. The prob-
formity to valid inferences. But this circle is a virtuous lem of induction is not a problem of demonstration but a
one. The point is that rules and particular inferences alike problem of defining the difference between valid and in-
are justified by being brought into agreement with each valid predictions.
other. A rule is amended if it yields an inference w e are This clears the air but leaves a lot to be done. As prin-
unwilling to accept; an inference is rejected if it violates a ciples of deductive inference, we have the familiar and
rule w e are unwilling to amend. The process of justifica- highly developed laws of logic; but there are available no
tion is the delicate one of making mutual adjustments such precisely stated and well-recognized principles of
between rules and accepted inferences; and in the agree- inductive inference. Mill's canons hardly rank with Aris-
ment achieved lies the only justification needed for totle's rules of the syllogism, let alone with Principia
either. W hasty reader might suppose that my insistence here upon
All this applies equally well to induction. An inductive identifying the problem of justification with a problem of descrip-
inference, too, is justified by conformity to general rules, tion is out of keeping with my parenthetical insistence in the pre-
and a general rule by conformity to accepted inductive ceding lecture that the goal of philosophy is something quite dif-
inferences. Predictions are justified if they conform to ferent from the mere description of ordinary or scientific pro-
valid canons of induction; and the canons are valid if they cedure. Let me repeat that the point urged there was that the
organization of the explanatory account need not reflect the man-
accurately codify accepted inductive practice. ner or order in which predicates are adopted in practice. It surely
A result of such analysis is that we can stop plaguing must describe practice, however, in the sense that the extensions
ourselves with certain spurious questions about induction. of predicates as explicated must conform in certain ways to the
W e no longer demand an explanation for guarantees that extensions of the same predicates as applied in practice. Hume's
we do not have, or seek keys to knowledge that we can- account is a description in just this sense. For it is an attempt to set
not obtain. It dawns upon us that the traditional smug forth the circumstances under which those inductive judgments
are made that are normally accepted as valid; and to do that is to
insistence upon a hard-and-fast line between justifying state necessary and sufficient conditions for, and thus to define,
induction and describing ordinary inductive practice dis- valid induction. What I am maintaining above is that the prob-
torts the problem. And we owe belated apologies to lem of justifying induction is not something over and above the
Hume. For in dealing with the question how normally problem of describing or defining valid induction.
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION
I
:
Mathematics. Elaborate and valuable treatises on proba- usually so considered. A definition may modify as well as
bility usually leave certain fundamental questions un- extend ordinary usage.3
touched. Only in very recent years has there been any Some pioneer work on the problem of defining con-
eqplicit and systematic work upon what I call the con- firmation o r valid induction has been done by Professor
structive task of confirmation theory. H e m ~ e l .Let
~ me remind you briefly of a few of his
results. Just as deductive logic is concerned primarily
with a relation between statements-namely the conse-
3. T h e Constructive Task of Confirnlation Theory
quence relation-that is independent of their truth or
T h e task of formulating rules that define the difference falsity, so inductive logic as Hempel conceives it is con-
between valid and invalid inductive inferences is much cerned primarily with a comparable relation of confirma-
like the task of defining any term. with an established tion between statements. Thus the problem is to define
usage. If we set out to define the term "tree", we try to the relation that obtains between any statement S1 and
compose out of already understood words an expression another Sz if and only if S1 may properly be said to con-
that will apply to the familiar objects that standard usage firm SZin any degree.
calls trees, and that will not apply to objects that standard With the question so stated, the first step seems obvi-
usage refuses to call trees. A proposal that plainly violates ous. Does not induction proceed in just the opposite
either condition is rejected; while a definition that meets direction from deduction? Surely some of the evidence-
these tests may be adopted and used to decide cases that statements that inductively support a general hypothesis
are not already settled by actual usage. Thus the interplay are consequences of it. Since the consequence relation is
we observed between rules of induction and particular already well defined by deductive logic, will we not be
inductive inferences is simply an instance of this charac- on firm ground in saying that confirmation embraces the
teristic dual adjustment between definition and usage, converse relation? T h e laws of deduction in reverse will
whereby the usage informs the definition,, which in turn then be among the laws of induction.
guides extension of the usage. Let's see where this leads us. W e naturally assume fur-
O f course this adjustment is a more complex matter
than I have indicated. Sometimes, in the interest of con- 3 For a fuller discussion of definition in general see Chapter I of
venience or theoretical utility, we deliberately permit a The Structure of Appearance.
definition to run counter to clear mandates of common The basic article is 'A Purely Syntactical Definition of Con-
usage. W e accept a definition of "fish" that excludes firmation', cited in Note 1.10. A much less technical account
is given in 'Studies in the Logic of Confirmation', Mind, n.s.,
whales. Similarly we may decide to deny the term "valid vol. 54 ( 1 9 4 5 ) pp.
~ 1-26 and 97-121. Later work by Hempel and
induction" to some inductive inferences that are com- others on defining degree of confirmation does not concern us
monly considered valid, or apply the term to others not here.
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION .
ther that whatever confirms a given statement confirms 8497 is a prime number and the other side of the moon is flat
also whatever follows from that ~tatement.~ But if we . and Elizabeth the First was crowned on a Tuesday.
combine this assumption with our proposed principle, we T o show that any one of the three component statements
get the embarrassing result that every statement confirms is true is to support the conjunction by reducing the net
every other. Surprising as it may be that such innocent undetermined claim. But support6 of this kind is not con-
beginnings lead to such an intolerable conclusion, the firmation; for establishment of one component endows
proof is very easy. Start with any statement SI. It is a the whole statement with no credibility that is transmit-
consequence of, and so by our present criterion confirms, ted to other component statements. Confirmation of a
the conjunction of S1 and any statement whatsoever- hypothesis occurs only when an instance imparts to the
call it S2. But the confirmed conjunction, Sl*S2,of course hypothesis some credibility that is conveyed to other
has S2 as a consequence. Thus every statement confirms instances. Appraisal of hypotheses, indeed, is incidental
all statements. to prediction, to the judgment of new cases on the basis
The fault lies in careless formulation of our first pro- of old ones.
posal. While some statements that confirm a general hy- Our formula thus needs tightening. This is readily
pothesis are consequences of it, not all its consequences accomplished, as Hempel points out, if we observe that
confirm it. This may not be immediately evident; for a hypothesis is genuinely confirmed only by a state-
indeed we do in some sense furnish support for a state- ment that is an instance of it in the special sense of
ment when we establish one of its consequences. W e entailing not the hypothesis itself but its relativization or
settle one of the questions about it. Consider the hetero- restriction to the class of entities mentioned by that
geneous conjunction: statement. The relativization of a general hypothesis to
a class results from restricting the range of its uni-
versal and existential quantifiers to the members of that
6 1 am not here asserting that this is an indispensable require- class. Less technically, what the hypothesis says of all
ment upon a definition of confirmation. Since our commonsense things the evidence statement says of one thing (or
assumptions taken in combination quickly lead us to absurd con-
clusions, some of these assumptions have to be dropped; and dif-
ferent theorists may make different decisions about which to drop 6 Any hypothesis is 'supported' by its own positive instances;
and which to preserve. Hempel gives up the converse conse- but support-or better, direct factual support-is only one factor
quence condition, while Carnap (Logical Foundations of Proba- in confirmation. This factor has been separately studied by John
bility, Chicago and London, 1950, pp. 474-6) drops both the G. Kemeny and Paul Oppenheim in 'Degree of Factual Support',
consequence condition and the converse consequence condition. Philosophy of Science, vol. 19 (1952)~pp. 307-24. AS will appear
Such differences of detail between different treatments of con- presently, my concern in these lectures is primarily with certain
firmation do not affect the central points I am making in this other important factors in confirmation, some of them quite gen-
lecture. erally neglected.
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION

of one pair or other wad of things). This obviously covers either black or a raven. In the light of this confirmation of
the confirnmation of the conductivity of all copper by the the hypothesis that there are no ravens, it is no longer sur-
conductivity of a given piece; and it excludes confirmation prising that under the artificial restrictions of the example,
of our heterogeneous conjunction b y any of its conmpo- the hypothesis that all ravens are black is also confirmed.
nents. ,4nd, when taken together with the principle that And the prospects for indoor ornithology vanish when we
what confirms a statement confirnms all its conseqnences, notice that under these same conditions, the contrary
this criterion does not yield the untoward conclusion that hypothesis that no ravens are black is equally well con-
every statenment confirnms every other. firmed.7
N e w difficulties promptly appear from other directions, O n the other hand, our definition does err in not forcing
however. One is the infamous paradox of .the ravens. T h e us to take into account all the stated evidence. T h e un-
statement that a given object, say this piece of paper, is happy results are readily illustrated. If two compatible evi-
neither black nor a raven confirms the hypothesis that all dence statements confirm two hypotheses, then naturally
non-black things are non-ravens. But this hypothesis is the conjunction of the evidence statements should confirm
logically equivalent to the hypothesis that all ravens i r e the conjunction of the hypo these^.^ Suppose our evidence
black. Hence we arrive at the unexpected conclusion that consists of the statements EI saying that a given thing b is
the statenment that a given object is neither black nor a black, and E2 saying that a second thing c is not black. By
raven confirms the hypothesis that all ravens are black. our present definition, EI confirms the hypothesis that
T h e prospect of being able to investigate ornithological everything is black, and E2 the hypothesis that everything
theories without going out in the rain is so attractive that is non-black. T h e conjunction of these perfectly compati-
we know there must be a catch in it. T h e trouble this time, ble evidence statements will then confirm the self-contra-
however, lies not in faulty definition, but in tacit and dictory hypothesis that everything is both black and non-
illicit reference to evidence not stated in our example. black. Simple as this anomaly is, it requires drastic modi-
Taken by itself, the statement that the given object is fication of our definition. What given evidence confirms
neither black nor a raven confirms the hypothesis that
everything that is not a raven is not black as well as the 7 An able and thorough exposition of this paragraph is given by
hypothesis that everything that is not black is not a raven. Israel Scheffler in his Anatomy of Inquiry, N e w York, 1963, pp.
W e tend to ignore the former hypothesis because we 28-1.

know it to be false from abundant other evidence-from 8The status of the conjunction condition is much like that
all the familiar things that are not ravens but are black. of the consequence condition-see Note 111.5. Although Carnap
drops the cbnjunction condition also (p. 394), he adopts for dif-
But we are required to assume that no such evidence is ferent reasons the requirement we find needed above: that the
available. Under this circumstance, even a much stronger total available evidence must always be taken into account (pp.
hypothesis is also obviously confirmed: that nothing is 2 1 1-13).
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION

is not what we arrive at by generalizing from separate its syntactical form. That a given piece of copper conducts
items of it, but-roughly speaking-what we arrive at by electricity increases the credibility of statements asserting
generalizing from the total stated evidence. The central that other pieces of copper conduct electricity, and thus
idea for an improved definition is that, within certain confirms the hypothesis that all copper conducts electric-
limitations, what is asserted to be true for the narrow ity. But the fact that a given 'man now in this room is a
universe of the evidence statements is confirmed for the third son does not increase the credibility of statements
whole universe of discourse. Thus if our evidence is El and asserting that other men now in this room are third sons,
Es neither the hypothesis that all things are black nor the and so does not confimm the hypothesis that all men now
hypothesis that all things are non-black is confirmed; for in this room are third sons. Yet in both cases our hypoth-
neither is true for the evidence-universe consisting of b esis is a generalization of the evidence statement. The
and c. Of course, much more careful formulation is difference is that in the former case the hypothesis is a law-
needed, since some statements that are true of the evi- like statement; while in the latter case, the hypothesis is a
dence-universe-such as that there is only one black thing merely contingent or accidental generality. Only a state-
-are obviously not confirmed for the whole universe. ment that is lawlike-regardless of its truth or falsity or its
These matters are taken care of by the studied formal defi- scientific importance-is capable of receiving confirma-
nition that Hempel develops on this basis; but we cannot tion from an instance of it; accidental statements are not.
and need not go into further detail here. Plainly, then, we must look for a way of distinguishidg
N o one supposes that the task of confirmation-theory lawlike from accidental statements.
has been completed. But the few steps I have reviewed- So long as what seems t o be needed is merely a way of
chosen partly for their bearing on what is to follow--show excluding a few odd and unwanted cases that are inadver-
how things move along once the problem of definition tently admitted by our definition of confirmation, the
displaces the problem of justification. Important and long- problem may not seem very hard or very pressing. W e
unnoticed questions are brought. to .light and answered; fully expect that minor defects will be found in our defi-
and we are encouraged to expect that the many remaining nition and that the necessary refinements will have to be
questions will in time yield to similar treatment. worked out patiently one after another. But some fur-
But our satisfaction is shortlived. N e w and serious ther examples will show that our present difficulty is of a
trouble begins to appear. much graver kind.
Suppose that all emeralds examined before a certain time
t are green? At time t, then, our observations support the
4. The New Riddle of Induction
Although the example used is different, the argument to fol-
Confirmation of a hypothesii by an instance depends low is substantially the same as that set forth in my note 'A
rather heavily upon features of the hypothesis other than Query on Confirmation', cited in Note 1.16.
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION

hypothesis that all emeralds are green; and this is in accord like hypotheses are genuinely confirmed; but we have no
with our definition of confirmation. O u r evidence state- criterion as yet for determining lawlikeness. And now we
ments assert that emerald a is green, that emerald b is green, see that without some such criterion, our definition not
and so on; and each confirms the general hypothesis that merely includes a few unwanted cilses, but is so con~pletely
all emeralds are green. So far, so good. ineffectual that it virtually escludes nothing. W e are left
N o w let me introduce another predicate less familiar once again with the intolerable result that anything con-
than "green". It is the predicate "grue" and it applies to firms anything. This difficulty cannot be set aside as an
all things examined before t just in case they are green but annoying detail to be taken care of in due course. It has
to other things just in case they are blue. Then at time t we t o be met before our definition will work at all.
have, for each evidence statement asserting that a given Nevertheless, the difficulty is often slighted because on
emerald is green, a parallel evidence statement asserting the surface there seem to be easy ways of dealing with it.
that that emerald is grue. And the statements that emerald Sometimes, for esample, the problein is thought to be
'
a is grue, that emerald b is grue, and so on, will each con- much like the parados of the ravens. \.Ye are here again,
firm the general hypothesis that all emeralds are grue. it is pointed out, making tacit and illegitimate use of in-
Thus according to our definition, the prediction that all formation outside the stated evidence: the information,
emeralds subsequently examined will be green and the for example, that different samples of one material are
prediction that all will be grue are alike confirmed by usually alike in conductivity, and the inforn~ationthat
evidence statements describing the same observations. But different men in a lecture audience are usually not alike in
if an emerald subsequently examined is grue, it is blue and the number of their older brothers. But while it is true that
hence not green. Thus although we are well aware which such information is being snluggled in, this does not by
of the two incompatible predictions is genuinely con- itself settle the matter as it settles the matter of the ravens.
firmed, they are equally well confirmed according to our There the point was that when the smuggled information
present definition. Moreover, it is clear that if we simply is forthrightly declared, its effect upon the confirmation
choose an appropriate predicate, then on the basis of these of the hypothesis in question is immediately and properly
same observations we shall have equal confirmation, by registered by the definition we are using. On the other
our definition, for any prediction whatever about other hand, if to our initial evidence we add statements con-
emeralds-or indeed about anything else.1° As in our cerning the conductivity of pieces of other materials or
earlier example, only the predictions subsumed under law- concerning the number of older brothers of members of
lo For instance, we shall have equal confirmation, by our present
emeroses are grue and hence the prediction that roses subsequently
definition, for the prediction that roses subsequently examined
esamined will be blue. The problem raised by such antecedents
will be blue. Let "emerose" apply just to emeralds examined be-
has been little noticed, but is no easier to meet than that raised
fore time t, and to roses examined later. Then all emeroses so far
by similarly perverse consequents. See further IV, 4 below.
examined are grue, and this confirms the hypothesis that all
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION

other lecture audiences, this will not in the least affect hypothesis that all iron conducts electricity enhances the
the confirmation, according to our definition, of the hy- lawlikeness of the hypothesis that all zirconium conducts
pothesis concerning copper or of that concerning this electricity, but does not similarly affectthe hypothesis that
lecture audience. Since our definition is insensitive to the all the objects on my desk conduct electricity. Wherein
bearing upon hypotheses of evidence so related to them, lies the difference? The first two hypotheses fall under the
even when the evidence is fully declared, the difficulty broader h y p o t h e s i ~ a l lit "H"-that every class of
about accidental hypotheses cannot be explained away on things of the same material is uniform in conductivity; the
the ground that such evidence is being surreptitiously first and third fall only under some such hypothesis as-
taken into account. call it "K"-that every class of things that are either all of
A more promising suggestion is to explain the matter in the same material or all on a desk is uniform in conductiv-
terms of the effect of this other evidence not directly upon ity. Clearly the important difference here is that evidence
the hypothesis in question but indirectly through other for a statement affirming that one of the classes covered by
hypotheses that are confirmed, according to our defini- H has the property in question increases the credibility of
tion, by such evidence. Our information about other ma- any statement affirming that another such class has this
terials does by our definition confirm such hypotheses as property; while nothing of the sort holds true with respect
that all pieces of iron conduct electricity, that no pieces of to K. But this is only to say that H is lawlike and K is not.
rubber do, and so on; and these hypotheses, the explanation W e are faced anew with the very problem we are trying
runs, impart to the hypothesis that a11 pieces of copper con- to solve: the problem of distinguishing between lawlike
duct electricity (and also to the hypothesis that none do) and accidental hypotheses.
the character of lawlikeness-that is, amenability to con- The most popular way of attacking the problem takes
firmation by direct positive instances when found. On the its cue from the fact that accidental hypotheses seem
other hand, our information about other lecture audiences typically to involve some spatial or temporal restriction, or
disconfirms many hypotheses to the effect that all the men reference to some particular individual. They seem to
in one audience are third sons, or that none are; and this concern the people in some particular room, or the objects
strips any character of lawlikeness from the hypothesis on some particular person's desk; while lawlike hypotheses
that all (or the hypothesis that none) of the men in this characteristically concern all ravens or all pieces of copper
audience are third sons. But clearly if this course is to be whatsoever. Complete generality is thus very often sup-
followed, the circumstances under which hypotheses are posed to be a sufficient condition of lawlikeness; but to
thus related to one another will have to be precisely arti- define this complete generality is by no means easy.
culated. Merely to require that the hypothesis contain no term
The problem, then, is to define the relevant way in naming, describing, or indicating a particular thing or
which such hypotheses must be alike. Evidence for the location will obviously not be enough. The troublesome
THE NEW RIDDLE OF INDUCTION a THE NEW RIDDLE OF INDUCTION

hypothesis that all emeralds are grue contains no such predicate is then conceived either as one that is equivalent
term; and where such a term does occur, as in hypotheses to some expression free of terms for specific individuals,
about men in this room, it can be suppressed in favor of or as one that is equivalent to no expression that contains
some predicate (short or long, new or old) that contains such a term; for this only raises again the difficulties just
no such term but applies only to exactly the same things. pointed out. T h e claim appeais to be rather that at least
One might think, then, of excluding not only hypotheses in the case of a simple enough predicate we can readily
that actually contain terms for specific individuals but determine by direct inspection of its meaning whether or
also all hypotheses that are equivalent to others that do not it is purely qualitative. But even aside from obscurities
contain such terms. But, as we have just seen, to exclude in the notion of 'the meaning' of a predicate, this claim
only hypotheses of which all equivalents contain such seems to me wrong. I simply do not know how to tell
terms is to exclude nothing. O n the other hand, to exclude whether a predicate is qualitative or positional, except
all hypotheses that have some equivalent containing such a perhaps by completely begging the question at issue and
term is to exclude everything; for even the hypothesis asking whether the predicate is 'well-behaved'-that is,
whether simple syntactically universal hypotheses apply-
All grass is green ing it are lawlike.
has as an equivalent This statement will not go unprotested. "Consider", it
will be argued, "the predicates 'blue' and 'green' and the
All grass in London or elsewhere is green. predicate 'grue' introduced earlier, and also the predicate
'bleen' that applies to emeralds examined before time t
T h e next step, therefore, has been to consider ruling just in case they are blue and to other emeralds just in
out predicates of certain kinds. A syntactically universal case they are green. Surely it is clear", the argument runs,
hypothesis is lawlike, the proposal runs, if its predicates "that the first two are purely qualitative and the second
are 'purely qualitative' or 'non-positional'." This will two are not; for the meaning of each of the latter two
obviously accomplish nothing if a purely qualitative plainly involves reference to'a specific temporal position."
T o this I reply that indeed I do recognize the first two as
Carnap took this course in his paper 'On the Application of well-behaved predicates admissible in lawlike hypotheses,
Inductive Logic', Philosophy and Phenomenological Research, and the second two as ill-behaved predicates. But the
VOI. 8 ( 1947), pp. 133-47, which is in part a reply to my 'A Query argument that the former but not the latter are purely
on Confirmation', cited in Note 1.16. T h e discussion was con-
tinued in my note 'On Infirmities of Confirmation Theory',
qualitative seems to me quite unsound. True enough, if we
Philosophy and Phenomenological Research, vol. 8 (1947)~pp. start with "blue" and "green", then "grue" and "bleen"
149-5 1; and in Carnap's 'Reply to Nelson Goodman', same jour- will be explained in terms of "blue" and "green" and a
nal, same volume, pp. 461-2. temporal term. But equally truly, if we start with "grue"
THE NEW RIDDLE OF INDUrnION THE NEW RIDDLE OF INDUCTION

and "bleen", then "blue" and "green" will be explained of confirmation. It is this problem that I call the new
in terms of digrue" and "bleen" and a temporal term; riddle of induction.
66
green", for example, applies to emeralds examined before
time t just in case they are grue, and to other emeralds J. The Pervasive Problem of Projection
just in case they are bleen. Thus qualitativeness is an en-
tirely relative matter and does not by itself establish any At the beginning of this lecture, I expressed the opinion
dichotomy of predicates. This relativity seems to be com- that the problem of induction is still unsolved, but that the
pletely overlooked by those who contend that the quali- difficulties that face us today are not the old ones; and I
tative character of a predicate is a criterion for its good have tried to outline the changes that have taken place.
behavior. The problem of justifying induction has been displaced
Of course, one may ask why we need worry about such by the problem of defining confirmation, and our work
unfamiliar predicates as "grue" or about accidental hy- upon this has left us with the residual problem of dis-
potheses in general, since we are unlikely to use them in tinguishing between confirmable and non-confirmable
making predictions. If our definition works for such hypotheses. One might say roughly that the first question
hypotheses as are normally employed, isn't that all we was "Why does a positive instance of a hypothesis give
need? In a sense, yes; but only in the sense that we need any grounds for predicting further instances?"; that the
no definition, no theory of induction, and no philosophy newer question was "What is a positive instance of a
of knowledge at all. W e get along well enough without hypothesis?"; and that the crucial remaining question is
them in daily life and in scientific research. But if we seek "What hypotheses are confirmed by their positive in-
a theory at all, we cannot excuse gross anomalies resulting stances?"
from a proposed theory by pleading that we can avoid The vast amount of effort expended on the problem of
them in practice. The odd cases we have been consider- induction in modern times has thus altered our afflictions
ing are clinically pure cases that, though seldom en- but hardly relieved them. The original difficulty about
countered in practice, nevertheless display to best induction arose from the recognition that anything may
advantage the symptoms of a widespread and destructive follow upon anything. Then, in attempting to define con-
malady. firmation in terms of the converse of the consequence
W e have so far neither any answer nor any promising relation, we found ourselves with the distressingly similar
clue to an answer to the question what distinguishes law- difficulty that our definition would make any statement
like or confirmable hypotheses from accidental or non- confirm any other. And now, after modifying our defini-
confirmable ones; and what may at first have seemed a tion drastically, we still get the old devastating result that
minor technical difficulty has taken on the stature of a any statement will confirm any statement. Until we find a
major obstacle to the development of a satisfactory theory way of exercising some control over the hypotheses to be
THE NEW RIDDLE OF INDUCTION THE NEW RIDDLE OF INDUCTION *

admitted, our definition makes no distinction whatsoever confirmed by statements related to it in the prescribed way
between valid and invalid inductive inferences. provided it is so confirmed. This is a good deal like having
T h e real inadequacy of Hume's account lay not in his a theory that tells us that the area of a plane figure is one-
descriptive approach but in the imprecision of his descrip- half the base times the altitude, without telling us for what
tion. Regularities in experience, according to him, give figures this holds. W e musr son~ehowfind a way of dis-
rise to habits of expectation; and thus it is predictions tinguishing lawlike hypotheses, to which our definition of
conforming to past regularities that are normal or valid. confirmation applies, from accidental hypotheses, to
But Hume overlooks the fact that some regularities do and which it does not.
some do not establish such habits; that predictions based Today I have been speaking solely of the problen~of
on some regularities are valid while predictions based on induction, but what has been said applies equally to the
other regularities are not. Every word you have heard me more general problem of projection. As pointed out
say has occurred prior to the final sentence of this lecture; earlier, the problem of prediction from past to future cases
but that does not, I hope, create any expectation that is but a narrower version of the problem of projecting
every word you will hear me say will be prior to that sen- from any set of cases to others. W e saw that a whole clus-
tence. Again, consider our case of emeralds. All those ter of troublesome problems concerning dispositions and
examined before time t are green; and this leads us to possibility can be reduced to this problem of projection.
expect, and confirms the prediction, that the next one That is why the new riddle of induction, which is more
will be green. But also, all those examined are grue; and broadly the problem of distinguishing between projectible
this does nor lead us ro expect, and does not confirm the and non-projectible hypotheses, is as important as it is
prediction, that the next one will be grue. Regularity in exasperating.
greenness confirms the prediction of further cases; regu- Our failures teach us, I think, that lawlike or project-
larity in grueness does not. T o say that valid predictions ible hypotheses cannot be distinguished on any merely
are those based on past regularities, without being able to syntactical grounds or even on the ground that these hy-
say which regularities, is thus quite pointless. Regularities potheses are somehow purely general in meaning. Our
are where you find them, and you can find them any- only hope lies in re-examining the problem once more
where. As we have seen, Hume's failure to recognize and and looking for some new approach. This will be my
deal with this problem has been shared even by his most course in the final lecture.
recent successors.
A s a result, what we have in current confirmation
theory is a definition that is adequate for certain cases that
so far can be described only as those for which it is ade-
quate. The theory works where it works.'^ hypothesis is
"The New
Riddle of
Induction."
Chapter III of
Nelson
Goodman
(1983): Fact,
Fiction, and
Forecast.
Cambridge,
MA: Harvard
University
Press, 59-83.

Un humilde tributo a nuestros maestros, porque enseñan a pesar de la


adversidad.

Traducción: Joel Tucídides Madrigal


Bailón, Licenciado en Filosofía, México
Distrito Federal. Universidad Autónoma
Metropolitana Unidad Iztapalapa
(tucidides2000@yahoo.com.mx). Para
educar y sin fines de lucro. Se permite la
reproducción total o parcial citando al
autor original.

Para Angélica, porque su cariño e


inteligencia son mi fuerza en la
adversidad.

Para Carmina Itzel Ruíz Madrigal, porque


su sonrisa es la prueba del buen futuro que
todavía está por venir.