Novel Prediction

Philosophy of Science

Muhammad Sajeer Bukhari

Chapter 14

Novel Prediction

Introduction to Novel Prediction

We recall the ongoing debate surrounding the confirmation of scientific theories, particularly focusing on what constitutes compelling evidence for a theory. At its core, a theory's validity hinges on its ability to align with observable data. When a theory predicts phenomena that have not yet been observed, this is termed as a novel prediction. In contrast, when a theory is designed to explain phenomena already known, it accommodates existing data. This chapter delves into the concept of novel prediction in the philosophy of science, exploring its significance in theory confirmation and its implications for understanding scientific progress.

Historical Examples

Throughout the history of science, several cases illustrate the power of novel predictions in confirming theories. Dmitri Mendeleev's periodic table, for instance, successfully predicted the properties of germanium, gallium, and scandium based on their atomic weights, which bolstered support for his theory. Similarly, Newtonian mechanics gained significant credibility after accurately predicting the return of Halley's Comet. Einstein's general relativity made waves with its prediction of the gravitational deflection of light by the sun. Fresnel's wave theory of light, notable for predicting the Arago spot—a bright spot at the center of a shadow—also stands as a testament to the potency of novel predictions in theory validation.

Views on Novel Prediction

In philosophical discourse, three primary views emerge regarding the role of novel predictions in theory confirmation:

 
1. Strong Predictivism : This perspective asserts that evidence only confirms a theory when it involves a novel prediction. Accommodation of existing data holds no confirmatory power in this framework.

2. Predictivism : While accommodating existing data is acknowledged as having some confirmatory power, novel predictions are considered superior evidence, all else being equal.

3. Accommodationism : Here, novel predictions and accommodation are considered equally valid. What matters is that a theory accurately entails the correct observations, regardless of whether they were already known or predicted anew.

Types of Novelty

It's essential to distinguish between types of novelty in predictions:

- Temporal Novelty : When evidence was not known at the time of theory construction. Examples include Fresnel's prediction of the Arago spot.

- Use Novelty : When evidence was not used in constructing the theory. This form of novelty often holds greater weight in confirmation discussions, as it suggests a theory's predictive power beyond tailored accommodations.

Philosophical Implications

The debate over novel predictions resonates deeply with discussions in scientific realism. Proponents argue that the success of theories in making accurate novel predictions supports the realism view—that our best scientific theories provide increasingly accurate descriptions of the world. Critics, however, challenge this view, questioning whether novel predictions truly necessitate a realist interpretation of scientific theories.

Arguments for Predictivism

The "no coincidence" argument is a prominent defense of predictivism. It posits that when a theory makes a correct novel prediction, the most plausible explanation is that the theory accurately describes the world, rather than the prediction being a mere coincidence. This argument draws parallels to the "no miracle" argument in scientific realism, asserting that successful predictions strongly suggest the truth or approximate truth of a theory.

Exploring the Design Hypothesis

The debate between truth and design hypotheses deepens when we consider their implications for both novel predictions and accommodation of evidence. In the case of accommodation, where a theory is constructed to fit known data, the design hypothesis asserts that the theory was intentionally formulated to explain these observations. This implies that while the theory might be true, its success in accommodating known data is primarily attributed to its intentional construction. According to proponents of this view, the truth of the theory becomes redundant as an explanation for its success in accommodating existing evidence.

However, the principle of avoiding redundant explanations suggests that if the design hypothesis is accepted and uncontroversial, it adequately explains why a theory accommodates known evidence. In such cases, introducing the truth hypothesis alongside the design hypothesis would be unnecessary and redundant. The widespread acceptance of the design hypothesis in accommodating scenarios reinforces this argument, indicating that we should prioritize intentional design over truth when evaluating theories based on their accommodation of known data.

Novel Predictions and the Truth Hypothesis

Contrasting with accommodation, novel predictions present a different challenge. Here, theories propose phenomena that were not previously known or expected at the time of their formulation. Unlike accommodation, where the design hypothesis fits neatly, novel predictions do not align with intentional design. The theory was not constructed with the aim of predicting these new phenomena, making the design hypothesis irrelevant in explaining the theory's success.

Moreover, in cases of novel predictions, proponents argue that truth serves as a more compelling explanation than coincidence. The "no coincidence" argument posits that if a theory successfully predicts a novel phenomenon, the most plausible explanation is that the theory accurately describes the underlying reality. This aligns closely with the principles of scientific realism, which holds that our best theories provide increasingly accurate depictions of the world.

Eric Barnes' Critique and Alternative Views

Eric Barnes offers a critical perspective in his work "Neither Truth nor Empirical Adequacy Explain Novel Success." Barnes argues against the idea that the design hypothesis renders the truth hypothesis redundant. He suggests that the intentions of the theorist, as encapsulated in the design hypothesis, explain why a theory was formulated in a particular manner. This includes why certain data were chosen to accommodate.

Barnes contends that the truth hypothesis, on the other hand, explains why the theory accurately reflects observed phenomena. It clarifies how the theory's claims correspond to empirical reality, which the design hypothesis alone cannot fully account for. Thus, Barnes challenges the notion that the design hypothesis negates the need for truth in explaining the success of a theory, particularly in cases of accommodation.

Patrick Maher's Illustrative Example

Patrick Maher's example further illustrates the distinction between prediction and accommodation. Maher presents scenarios involving coin flips where one scenario involves predicting an additional flip based on known data (accommodation), and another scenario involves predicting flips without prior knowledge (novel prediction). He argues that the success of novel predictions provides stronger confirmation of a theory's reliability compared to accommodation, where success may appear more arbitrary or coincidental.

Critiques and Responses

However, critiques by Mark Lang suggest that apparent superiority of novel predictions over accommodation may not solely stem from their predictive nature but could involve other factors, such as the coherence of the theory's claims across different cases. Lang's perspective highlights that the distinction between prediction and accommodation may not always be straightforward and may depend on the coherence and logical structure of the theory itself.

The Challenge of Arbitrary Conjunctions

Mark Lang's critique of Patrick Maher's distinction between prediction and accommodation centers on the concept of arbitrary conjunctions. Lang argues that the perceived superiority of novel predictions over accommodation may not solely hinge on their predictive nature but on whether the theory in question avoids becoming an arbitrary conjunction. A theory becomes an arbitrary conjunction when its various claims lack cohesion or logical connection with each other. Lang illustrates this with an example: a theory stating that protons contain two up quarks, all animals share a common ancestor, and Captain Beefheart was born in 1941.

Clearly, such a theory lacks coherence; its components appear unrelated to each other. While each claim could potentially be confirmed individually, their conjunction does not derive from a unified explanatory framework. Lang contends that theories resembling arbitrary conjunctions typically arise after data collection, as they are designed to accommodate existing evidence rather than predict new phenomena.

Maher's Illustration and Lang's Counterexample

In Maher's example of coin flips, the distinction between prediction and accommodation becomes apparent. In the scenario where Frank predicts the 100th flip based on known data (accommodation), his theory appears arbitrary because it merely extends a sequence without a deeper explanatory basis. Conversely, in Vincent's case, where he predicts the entire sequence without prior knowledge (novel prediction), the successful prediction suggests a deeper understanding or method underlying the theory's formulation.

Lang further explores this distinction by contrasting different scenarios of coin flips. For instance, if the sequence alternates predictably, both Frank and Vincent might be able to predict the 100th flip accurately, albeit for different reasons. In such cases, the predictability of outcomes diminishes the importance of whether the prediction was novel or accommodative, focusing instead on the explanatory power or method behind the prediction.

Critique and Real-World Application

Lang's critique challenges the straightforward application of predictivism by suggesting that the crucial factor may not always be whether a prediction is novel or accommodative but whether the theory avoids becoming an arbitrary conjunction. Predictive success that stems from a clear method or explanatory framework may offer more compelling confirmation of a theory's reliability than mere novelty.

Furthermore, Lang's argument prompts consideration of how theories are constructed and tested in real-world scientific practice. The ability of a theory to make falsifiable predictions, whether novel or accommodative, remains a cornerstone of scientific methodology. Falsifiability ensures that theories face rigorous testing against new data, thereby strengthening or refuting their validity based on empirical evidence.

Falsifiability and Theory Evaluation

The concept of falsifiability, championed by philosophers like Karl Popper, emphasizes that scientific theories must make predictions that could potentially disprove them. Novel predictions, by proposing phenomena not previously accounted for, offer stringent tests of a theory's validity. In contrast, accommodation of existing evidence, while necessary for theory-building, may not provide the same rigorous testing grounds.

Accommodation, according to Popper, does not inherently confirm or strengthen a theory because it does not subject the theory to the risk of being falsified by new evidence. A theory designed to fit existing data may appear robust but lacks the critical element of being tested against the unpredictability of new observations.

Responses and Counterarguments

Critics of falsifiability's centrality in theory evaluation argue that its importance lies not just in potentially disproving theories but in guiding scientific inquiry and prediction. The ability of a theory to generate expectations about future observations, regardless of whether they are novel or accommodative, remains essential in advancing scientific understanding and application.

Defending Accommodationism

Accommodationism offers several robust defenses against predictivism, each challenging the notion that only novel predictions hold significant weight in theory evaluation.

1. Falsifiability and Accommodation

One counterargument posits that the emphasis on falsifiability as a criterion for theory evaluation may overlook the broader utility of theories in scientific practice. Falsifiability suggests that theories should be vulnerable to empirical testing, which enhances their credibility by ruling out potential outcomes. However, theories also serve a crucial role in guiding expectations and predictions about future observations. For instance, when planning a space mission to Pluto, scientists rely on gravitational theories to predict the probe's trajectory based on interactions with planetary bodies.

If a theory becomes unfalsifiable, proponents argue, it may resort to ad hoc hypotheses to shield itself from refutation, thereby compromising its predictive power. The specificity and reliability of predictions, such as those provided by general relativity in predicting gravitational effects precisely, underscore the importance of falsifiability in ensuring that theories provide meaningful guidance and explanatory power.

2. Accommodation and Explanation

Another defense of accommodationism focuses on the explanatory power of theories in accommodating evidence. A theory that accommodates evidence effectively explains why certain phenomena occur as observed, thereby ruling out alternative scenarios. When theories incorporate ad hoc hypotheses to accommodate new data without enhancing explanatory coherence, they fail to genuinely accommodate the evidence. This perspective suggests that an unfalsifiable theory loses its ability to provide robust explanations, thereby undermining its scientific value.

3. Bayesian Confirmation Theory

A third argument appeals to Bayesian epistemology, which assesses how evidence updates our beliefs about the likelihood of a theory's truth. According to Bayesianism, evidence confirms a theory if it increases the probability of the theory being true. Novel predictions, by introducing new, potentially surprising evidence, inherently raise the probability of the theory if confirmed. In contrast, old evidence already known to be true (accommodation) does not alter the probability of the theory significantly because it merely reaffirms what is already assumed.

Bayesianism highlights that the strength of confirmation depends on how much the evidence shifts our prior beliefs about the theory's truthfulness. Accommodation of old evidence, where the probability of the evidence is already high (often 1), does not enhance our confidence in the theory's validity as substantially as novel predictions might.

Integrating Perspectives

These defenses suggest that while predictivism emphasizes the predictive and falsifiable aspects of theories, accommodationism highlights their explanatory depth and ability to integrate observed phenomena comprehensively. Both approaches contribute valuable insights to understanding how theories are evaluated in scientific practice. Predictivism underscores the importance of challenging theories through novel predictions, while accommodationism emphasizes the explanatory coherence and utility of theories in explaining a wide range of observations.

Philosophical Reflections

In conclusion, the debate between predictivism and accommodationism reflects deeper philosophical questions about the nature of scientific theories and their role in understanding the world. While predictivism draws strength from historical examples of successful novel predictions, accommodationism defends the broader utility of theories in guiding scientific inquiry and forming coherent explanations of observed phenomena.

The distinction between the context of discovery and the context of justification, as traditionally discussed in philosophy of science, underscores that while the process of theory construction may be influenced by various factors, including psychological biases or historical contingencies, what ultimately matters in evaluating a theory's validity is its ability to explain and predict empirical phenomena effectively.

By integrating these perspectives, philosophers continue to refine our understanding of how theories are validated and refined in scientific inquiry, highlighting the complex interplay between predictive power, explanatory coherence, and empirical testing in theory evaluation.

Competing Intuitions and Historical Arguments

Despite the intuitive appeal of predictivism, accommodationism challenges the notion that the process of theory discovery should significantly influence our acceptance of scientific theories. Predictivism asserts that how a theory is conceived and developed should impact its justification—a stance that appears counterintuitive to many, who argue that the empirical evidence and explanatory power of a theory should stand independently of its historical origins.

Context of Discovery vs. Justification

Philosophers of science traditionally distinguish between the context of discovery and the context of justification. The context of discovery concerns the non-rational, often idiosyncratic processes by which theories are conceived, which may involve intuition, serendipity, or even dream-like insights. In contrast, the context of justification focuses on the empirical evidence, logical consistency, and explanatory coherence that support a theory's validity.

Predictivism challenges this distinction by suggesting that the historical circumstances of a theory's discovery—such as the psychological state or biases of its proponents—should influence our assessment of its validity. This perspective implies that if the historical record were to reveal that a theory's development was less rigorous or based on incomplete data, our confidence in that theory should decrease accordingly.

The History of Science Argument

Accommodationism draws strength from historical analyses that question the primacy of novel predictions in theory acceptance. Stephen G. Brush's studies, such as his examination of quantum mechanics, general relativity, and Mendeleev's periodic law, argue that historical cases often prioritize explanatory power and the ability to resolve anomalies over novel predictions. For example, quantum mechanics gained acceptance not primarily through its novel predictions but because it unified disparate phenomena and provided a consistent framework for calculations.

Similarly, Mendeleev's periodic law was acclaimed for its ability to classify elements based on their properties and atomic weights, rather than solely on its few successful novel predictions of new elements. These historical cases suggest that scientists often value a theory's ability to accommodate existing data and provide robust explanations more than its capacity for novel prediction alone.

Philosophical Reflections on Predictivism

While historical analyses provide empirical support for accommodationism, they also raise philosophical questions about the normative stance of predictivism. Predictivism asserts that novel predictions inherently provide stronger confirmation of a theory, despite historical evidence suggesting otherwise. Critics argue that this discrepancy between theory and practice in scientific history highlights potential flaws in predictivist reasoning, suggesting that the behavior of scientists does not always align with normative claims about theory evaluation.

Behavior of Scientists and Normative Claims

The argument for accommodationism based on historical precedent examines how scientists have actually behaved in evaluating theories. It acknowledges that scientists are fallible and sometimes accept theories for reasons other than empirical evidence alone—such as rhetorical persuasion or personal biases. While predictivism posits a normative claim that novel predictions should carry greater weight in theory confirmation, the actual practices of scientists often diverge from this ideal.

Accommodationists argue that philosophy of science should reflect and clarify existing scientific standards rather than impose normative standards that are not followed in practice. They contend that if scientists generally do not prioritize novel predictions as strongly as predictivism suggests, then predictivism may not accurately reflect how scientific knowledge is actually constructed and justified.

The Role of Philosophy of Science

Philosophy of science, according to this view, aims to articulate and make explicit the implicit standards that scientists follow in evaluating theories. It seeks to describe the criteria by which theories are judged and confirmed without necessarily advocating for changes in those criteria. This perspective emphasizes understanding and clarifying the process of theory confirmation as it occurs, rather than prescribing how it should occur.

However, there remains debate over whether philosophy of science can and should advocate for changes in scientific practice based on normative claims. Some argue that philosophy of science has a role in critiquing and potentially reforming scientific standards if they are found to be inconsistent or inadequate.

Superiority Over Rivals Argument

A second argument for accommodationism is the superiority over rivals argument. This perspective contends that while novel predictions can highlight the predictive power of a theory, accommodations—especially successful ones—can demonstrate a theory's superiority over competing explanations more decisively.

For instance, Einstein's theory of general relativity accommodated the long-standing anomaly of Mercury's perihelion procession, which no other theory had successfully explained. This successful accommodation played a crucial role in establishing general relativity as the leading explanation, as it directly addressed and resolved a persistent anomaly that had defied prior explanations. In contrast, Einstein's novel prediction of gravitational light bending, while remarkable, did not have the same power to decisively eliminate competing theories since it was a new phenomenon without a history of failed explanations by other theories.

Conclusion

In conclusion, the debate between predictivism and accommodationism in theory confirmation remains rich with nuanced arguments and historical examples. Predictivism posits that novel predictions hold special epistemic value, suggesting that they provide stronger confirmation for theories compared to accommodation of data. On the other hand, accommodationism argues that while novel predictions are valuable, they are not inherently superior to accommodations, which can also enhance the credibility and robustness of a theory.

Throughout our exploration, we've seen how predictivists emphasize the unique evidential power of novel predictions to decisively test and confirm theories, often through the elimination of rival hypotheses. Meanwhile, accommodationists highlight the role of accommodations in showing the adaptability and explanatory power of theories, especially in resolving anomalies and integrating new data into existing frameworks.

Moreover, considerations such as the historical evidence of scientific practice, the reliability of data, and the broader epistemic virtues of theories contribute to the complexity of this debate. While predictivism appeals to the idealized view of how science should ideally operate, accommodationism grounds its arguments in the actual practices and successes of scientific inquiry.

Ultimately, whether one favors predictivism or accommodationism often depends on how one weighs the significance of novel predictions against the broader context of scientific methodology, explanatory power, and simplicity. Both positions offer valuable insights into how scientific theories gain acceptance and evolve over time, reflecting the dynamic nature of scientific knowledge and inquiry.