Reading Passage
The Revival of Traditional Deep-Water Aquaculture
For decades, industrial marine aquaculture has predominantly occupied shallow coastal bays and sheltered estuaries. While these nearshore operations offered convenient accessibility and lower capital expenditure, their high spatial concentration has wrought significant ecological damage. Dense aggregations of coastal fish pens frequently trigger localized eutrophication, as unconsumed nutrient feeds and concentrated metabolic waste deplete dissolved oxygen levels, forming hypoxic dead zones. Furthermore, the stagnant water dynamics of enclosed bays foster parasitic infestations, particularly sea lice, which routinely spill over into wild fish populations and necessitate extensive chemical treatments.
To resolve these compounding environmental crises, marine engineers and oceanographers are turning toward deep-water open-ocean aquaculture—often referred to as offshore mariculture. Operating in high-energy oceanic environments typically located more than three nautical miles from shore and at depths exceeding 50 meters, offshore installations leverage deep marine currents to disperse organic effluents rapidly. Continuous nutrient dilution prevents localized benthic accumulation, maintaining water quality and preserving seabed ecosystems. Moreover, the strong hydrodynamic forces and cooler thermal profiles of open ocean waters significantly suppress parasite lifecycles, virtually eliminating the need for prophylactic pharmaceutical interventions.
Engineering structures capable of withstanding the relentless kinetic force of the open sea presents unprecedented technical challenges. Traditional flexible surface pens crumble under severe wave action during storm events. Consequently, modern offshore operations utilize fully submersible rigid cages constructed from high-density polyethylene frameworks and copper-alloy mesh netting. These sophisticated enclosures feature automated variable-buoyancy systems. During calm sea conditions, the cages float near the surface to maximize natural light and gas exchange. However, when satellite weather monitoring detects severe ocean swells or hurricane conditions, automated ballast tanks flood with seawater, sinking the entire structure up to 40 meters below the surface—safely beneath the wave-action zone.
In addition to advanced cage mechanics, offshore mariculture relies heavily on autonomous remote management systems. Operating far from terrestrial infrastructure, these farms deploy floating power buoys equipped with solar panels, wind turbines, and wave-energy converters to achieve electrical autonomy. Integrated acoustic telemetry sensors continuously log water temperature, salinity, dissolved oxygen, and fish biomass metrics, transmitting real-time data back to land-based control centers via satellite links. Automated feeding cannons, guided by computerized computer-vision algorithms, disperse specialized floating feed pellets based on real-time fish appetite indicators, minimizing feed waste and optimizing feed conversion ratios.
Despite its vast potential to secure global seafood supplies sustainably, the economic feasibility of deep-water aquaculture remains constrained by substantial financial and legal hurdles. Building storm-resistant submersible cages and autonomous support vessels requires capital expenditures up to five times higher than conventional nearshore facilities. Furthermore, because offshore installations frequently operate beyond national territorial waters in exclusive economic zones (EEZs) or international high seas, vague regulatory frameworks and overlapping jurisdictional claims create administrative inertia, delaying corporate investment in large-scale offshore infrastructure.
The Mystery of the Antikythera Mechanism
For half a century, the function of the so-called Antikythera Mechanism remained an intense subject of debate among historians. Classical scholarship had long held that complex geared mechanisms—specifically differential gear trains—were entirely unknown to antiquity, emerging only during the European Renaissance in the 14th century. Consequently, several early scholars dismissed the device as anachronistic, suggesting it was an early modern astrolabe dropped on the shipwreck centuries after the original disaster. However, in the 1970s, British physicist Derek de Solla Price conducted groundbreaking gamma-ray and X-ray radiography examinations of the 82 surviving fragments, proving that the gears were indisputably embedded within the ancient corrosion layers of the Roman wreck.
Price’s initial findings revealed that the mechanism contained a precision train of at least 30 interlocking bronze gears, featuring triangular teeth cut with extraordinary mathematical precision. Rather than tracking time like a mechanical clock, the device was an analog computer designed to calculate and display astronomical positions. The front face featured a large dial with two concentric rings showing the Greek zodiac and the Egyptian calendar, while two back dials displayed complex astronomical cycles: the Metonic cycle (a 19-year period synchronizing solar years and lunar months) and the Saros cycle (a 223-month cycle used to predict solar and lunar eclipses).
The full extent of the device's sophistication was revealed in the 2000s by the Antikythera Mechanism Research Project, using high-resolution microfocus X-ray computed tomography (CT scanning) and polynomial texture mapping. The 3D scans uncovered hidden inscriptions—tiny Greek texts rendered in letters less than two millimeters high—functioning as a user manual built directly into the bronze casing. Most surprisingly, the scans revealed a unique "pin-and-slot" gear mechanism. By connecting two gear wheels with slightly offset axes via a sliding pin, the ancient Greek makers successfully modeled Kepler’s second law of planetary motion seventeen centuries before Johannes Kepler mathematically defined it: accounting for the moon's irregular variable speed as it moves closer to and farther from Earth during its elliptical orbit.
The technical sophistication of the Antikythera Mechanism raises fundamental historical questions regarding the loss and transmission of technological knowledge. Historical texts mention that great Hellenistic mathematicians, notably Archimedes of Syracuse and Posidonius of Rhodes, constructed mechanical planetariums or orreries capable of replicating planetary paths. Yet, following the decline of the Hellenistic world and the consolidation of the Roman Empire, the specialized knowledge required to manufacture such high-precision gearing appears to have vanished from the Western Mediterranean region.
Where did this technological lineage go? Historians now believe that the knowledge was preserved and refined in the Islamic world during the Middle Ages. Arab scholars preserved classical Greek texts and developed complex geared astrolabes and calendars, such as those designed by the 11th-century polymath Al-Biruni. Eventually, this mechanical heritage re-entered Western Europe through Islamic Iberia and trade routes, catalyzing the development of the mechanical astronomical clocks that began gracing European cathedrals during the early Renaissance. The Antikythera Mechanism thus stands as a singular, miraculous physical bridge between ancient Hellenistic brilliance and modern mechanical engineering.
The Epistemology of Artificial Intelligence in Scientific Discovery For centuries, the scientific method has operated under an epistemological framework anchored in human-interpretable hypothesis generation, empirical experimentation, and theoretical modeling. Traditional scientific discovery relies heavily on reductionism—breaking complex physical phenomena down into fundamental mechanistic laws expressed through symbolic mathematics or logical structures. However, the integration of deep artificial neural networks into basic scientific research is forcing a radical paradigm shift. From predicting complex protein foldings via systems like AlphaFold to discovering novel quantum materials, machine learning models are solving scientific problems that long resisted human analytical capabilities. Yet, this success introduces a profound philosophical debate regarding the nature of scientific explanation and understanding.
At the core of this debate lies the "black box" problem of deep learning architectures. Deep neural networks achieve high predictive accuracy by adjusting millions, or even billions, of internal parameters (weights and biases) across non-linear hidden layers during training. Unlike traditional physical models, which provide transparent, symbolic explanations grounded in causal principles (such as Newton's laws of motion or Maxwell's equations), deep learning systems generate predictions based on multidimensional statistical correlations that evade human cognitive comprehension. Consequently, while AI can predict what will happen in a complex biological or physical system with unprecedented precision, it frequently fails to explain why it happens in terms that human scientists can intuitively grasp.
This epistemological tension has split philosophers of science into two opposing camps: instrumentalists and explainability realists. Instrumentalists argue that predictive utility is the ultimate measure of scientific success. Under this pragmatist view, if an AI model accurately predicts molecular binding affinities or forecasts climate patterns better than any human-derived theory, demanding human-style explanatory narratives is an unnecessary anthropocentric limitation. They argue that nature is under no obligation to be simple enough for human brains to understand, and if machine learning uncovers non-linear, high-dimensional patterns beyond human comprehension, science should embrace these "opaque predictions" to accelerate technological progress.
Conversely, explainability realists contend that true science requires explanation, not merely prediction. They argue that without mechanistic understanding—knowing the underlying causal pathways governing a phenomenon—scientists cannot identify when a model is relying on spurious correlations or domain-specific artifacts. A neural network might achieve 99% accuracy in diagnosing diseases from medical images by secretly learning to identify the brand of X-ray scanner used in the training dataset rather than actual pathological biomarkers. Without explainable AI (XAI) frameworks that translate neural network parameters into interpretable causal models, science risks descending into a form of "digital alchemy," where researchers blindly trust predictive engines without understanding the principles driving their output.
To bridge this divide, computer scientists are developing hybrid systems that merge data-driven deep learning with symbolic AI—a domain known as neuro-symbolic AI. By embedding established physical laws, conservation principles, and mathematical symmetries directly into the architecture of neural networks, neuro-symbolic systems force algorithms to operate within the bounds of known physical laws. These constrained models not only require significantly less training data, but they also produce outputs that are inherently interpretable and physically consistent. Ultimately, the future of AI-driven scientific discovery may not lie in replacing human scientific reasoning, but in expanding it through hybrid systems that pair machine pattern recognition with human conceptual synthesis.
Questions
Questions 1–7
Do the following statements agree with the information given in Reading Passage 1?
In boxes 1–7 on your answer sheet, write:
TRUE if the statement agrees with the information
FALSE if the statement contradicts the information
NOT GIVEN if there is no information on this
Questions 8–13
Complete the summary below.
Choose NO MORE THAN TWO WORDS from the passage for each answer.
Engineering Innovations in Offshore Mariculture
Offshore mariculture overcomes environmental degradation by placing fish farms in high-energy marine environments. To survive harsh weather, modern operations use submersible cages made with frameworks of and copper-alloy mesh. When satellite monitoring indicates approaching storms, these structures flood their to lower the cages beneath dangerous wave motion. These remote installations achieve energy independence by utilizing floating buoys powered by , wind, and wave energy. Data regarding water conditions and fish mass is sent to shore via . To reduce financial losses from unused food, algorithms operate to distribute pellets efficiently. However, broad adoption is still hindered by high along with complex jurisdictional issues in international waters.
Questions 14–19
Reading Passage 2 has six paragraphs, labeled A–F.
Which paragraph contains the following information?
Write the correct letter, A–F, in boxes 14–19 on your answer sheet.
Questions 20–22
Choose THREE letters, A–F.
Write the correct letters in boxes 20–22 on your answer sheet.
Which THREE astronomical phenomena or calendar calculations were tracked using the dials and gears of the Antikythera Mechanism?
Questions 23–26
Complete the sentences below.
Choose NO MORE THAN TWO WORDS from the passage for each answer.
Questions 27–31
Choose the correct letter, A, B, C, or D.
Write the correct letter in boxes 27–31 on your answer sheet.
Questions 32–36
Do the following statements agree with the claims of the writer in Reading Passage 3?
Write YES, NO, or NOT GIVEN.
Questions 37–40
Complete the summary using the list of words, A–H, below.
Write the correct letter, A–H, in boxes 37–40 on your answer sheet.
The role of artificial intelligence in research has sparked an epistemological debate. Traditional science seeks mechanistic explanation, whereas deep learning relies on hidden parameters generating predictions through . Pragmatic instrumentalists prioritize , arguing that complex natural patterns need not conform to human cognitive limits. However, realists warn that ignoring causal pathways leads to based on unintended dataset artifacts. To resolve this, researchers are developing , which integrate fundamental physical principles directly into AI architectures to ensure interpretable, reliable results.
Reading Passage
The Revival of Traditional Deep-Water Aquaculture
For decades, industrial marine aquaculture has predominantly occupied shallow coastal bays and sheltered estuaries. While these nearshore operations offered convenient accessibility and lower capital expenditure, their high spatial concentration has wrought significant ecological damage. Dense aggregations of coastal fish pens frequently trigger localized eutrophication, as unconsumed nutrient feeds and concentrated metabolic waste deplete dissolved oxygen levels, forming hypoxic dead zones. Furthermore, the stagnant water dynamics of enclosed bays foster parasitic infestations, particularly sea lice, which routinely spill over into wild fish populations and necessitate extensive chemical treatments.
To resolve these compounding environmental crises, marine engineers and oceanographers are turning toward deep-water open-ocean aquaculture—often referred to as offshore mariculture. Operating in high-energy oceanic environments typically located more than three nautical miles from shore and at depths exceeding 50 meters, offshore installations leverage deep marine currents to disperse organic effluents rapidly. Continuous nutrient dilution prevents localized benthic accumulation, maintaining water quality and preserving seabed ecosystems. Moreover, the strong hydrodynamic forces and cooler thermal profiles of open ocean waters significantly suppress parasite lifecycles, virtually eliminating the need for prophylactic pharmaceutical interventions.
Engineering structures capable of withstanding the relentless kinetic force of the open sea presents unprecedented technical challenges. Traditional flexible surface pens crumble under severe wave action during storm events. Consequently, modern offshore operations utilize fully submersible rigid cages constructed from high-density polyethylene frameworks and copper-alloy mesh netting. These sophisticated enclosures feature automated variable-buoyancy systems. During calm sea conditions, the cages float near the surface to maximize natural light and gas exchange. However, when satellite weather monitoring detects severe ocean swells or hurricane conditions, automated ballast tanks flood with seawater, sinking the entire structure up to 40 meters below the surface—safely beneath the wave-action zone.
In addition to advanced cage mechanics, offshore mariculture relies heavily on autonomous remote management systems. Operating far from terrestrial infrastructure, these farms deploy floating power buoys equipped with solar panels, wind turbines, and wave-energy converters to achieve electrical autonomy. Integrated acoustic telemetry sensors continuously log water temperature, salinity, dissolved oxygen, and fish biomass metrics, transmitting real-time data back to land-based control centers via satellite links. Automated feeding cannons, guided by computerized computer-vision algorithms, disperse specialized floating feed pellets based on real-time fish appetite indicators, minimizing feed waste and optimizing feed conversion ratios.
Despite its vast potential to secure global seafood supplies sustainably, the economic feasibility of deep-water aquaculture remains constrained by substantial financial and legal hurdles. Building storm-resistant submersible cages and autonomous support vessels requires capital expenditures up to five times higher than conventional nearshore facilities. Furthermore, because offshore installations frequently operate beyond national territorial waters in exclusive economic zones (EEZs) or international high seas, vague regulatory frameworks and overlapping jurisdictional claims create administrative inertia, delaying corporate investment in large-scale offshore infrastructure.
The Mystery of the Antikythera Mechanism
For half a century, the function of the so-called Antikythera Mechanism remained an intense subject of debate among historians. Classical scholarship had long held that complex geared mechanisms—specifically differential gear trains—were entirely unknown to antiquity, emerging only during the European Renaissance in the 14th century. Consequently, several early scholars dismissed the device as anachronistic, suggesting it was an early modern astrolabe dropped on the shipwreck centuries after the original disaster. However, in the 1970s, British physicist Derek de Solla Price conducted groundbreaking gamma-ray and X-ray radiography examinations of the 82 surviving fragments, proving that the gears were indisputably embedded within the ancient corrosion layers of the Roman wreck.
Price’s initial findings revealed that the mechanism contained a precision train of at least 30 interlocking bronze gears, featuring triangular teeth cut with extraordinary mathematical precision. Rather than tracking time like a mechanical clock, the device was an analog computer designed to calculate and display astronomical positions. The front face featured a large dial with two concentric rings showing the Greek zodiac and the Egyptian calendar, while two back dials displayed complex astronomical cycles: the Metonic cycle (a 19-year period synchronizing solar years and lunar months) and the Saros cycle (a 223-month cycle used to predict solar and lunar eclipses).
The full extent of the device's sophistication was revealed in the 2000s by the Antikythera Mechanism Research Project, using high-resolution microfocus X-ray computed tomography (CT scanning) and polynomial texture mapping. The 3D scans uncovered hidden inscriptions—tiny Greek texts rendered in letters less than two millimeters high—functioning as a user manual built directly into the bronze casing. Most surprisingly, the scans revealed a unique "pin-and-slot" gear mechanism. By connecting two gear wheels with slightly offset axes via a sliding pin, the ancient Greek makers successfully modeled Kepler’s second law of planetary motion seventeen centuries before Johannes Kepler mathematically defined it: accounting for the moon's irregular variable speed as it moves closer to and farther from Earth during its elliptical orbit.
The technical sophistication of the Antikythera Mechanism raises fundamental historical questions regarding the loss and transmission of technological knowledge. Historical texts mention that great Hellenistic mathematicians, notably Archimedes of Syracuse and Posidonius of Rhodes, constructed mechanical planetariums or orreries capable of replicating planetary paths. Yet, following the decline of the Hellenistic world and the consolidation of the Roman Empire, the specialized knowledge required to manufacture such high-precision gearing appears to have vanished from the Western Mediterranean region.
Where did this technological lineage go? Historians now believe that the knowledge was preserved and refined in the Islamic world during the Middle Ages. Arab scholars preserved classical Greek texts and developed complex geared astrolabes and calendars, such as those designed by the 11th-century polymath Al-Biruni. Eventually, this mechanical heritage re-entered Western Europe through Islamic Iberia and trade routes, catalyzing the development of the mechanical astronomical clocks that began gracing European cathedrals during the early Renaissance. The Antikythera Mechanism thus stands as a singular, miraculous physical bridge between ancient Hellenistic brilliance and modern mechanical engineering.
The Epistemology of Artificial Intelligence in Scientific Discovery For centuries, the scientific method has operated under an epistemological framework anchored in human-interpretable hypothesis generation, empirical experimentation, and theoretical modeling. Traditional scientific discovery relies heavily on reductionism—breaking complex physical phenomena down into fundamental mechanistic laws expressed through symbolic mathematics or logical structures. However, the integration of deep artificial neural networks into basic scientific research is forcing a radical paradigm shift. From predicting complex protein foldings via systems like AlphaFold to discovering novel quantum materials, machine learning models are solving scientific problems that long resisted human analytical capabilities. Yet, this success introduces a profound philosophical debate regarding the nature of scientific explanation and understanding.
At the core of this debate lies the "black box" problem of deep learning architectures. Deep neural networks achieve high predictive accuracy by adjusting millions, or even billions, of internal parameters (weights and biases) across non-linear hidden layers during training. Unlike traditional physical models, which provide transparent, symbolic explanations grounded in causal principles (such as Newton's laws of motion or Maxwell's equations), deep learning systems generate predictions based on multidimensional statistical correlations that evade human cognitive comprehension. Consequently, while AI can predict what will happen in a complex biological or physical system with unprecedented precision, it frequently fails to explain why it happens in terms that human scientists can intuitively grasp.
This epistemological tension has split philosophers of science into two opposing camps: instrumentalists and explainability realists. Instrumentalists argue that predictive utility is the ultimate measure of scientific success. Under this pragmatist view, if an AI model accurately predicts molecular binding affinities or forecasts climate patterns better than any human-derived theory, demanding human-style explanatory narratives is an unnecessary anthropocentric limitation. They argue that nature is under no obligation to be simple enough for human brains to understand, and if machine learning uncovers non-linear, high-dimensional patterns beyond human comprehension, science should embrace these "opaque predictions" to accelerate technological progress.
Conversely, explainability realists contend that true science requires explanation, not merely prediction. They argue that without mechanistic understanding—knowing the underlying causal pathways governing a phenomenon—scientists cannot identify when a model is relying on spurious correlations or domain-specific artifacts. A neural network might achieve 99% accuracy in diagnosing diseases from medical images by secretly learning to identify the brand of X-ray scanner used in the training dataset rather than actual pathological biomarkers. Without explainable AI (XAI) frameworks that translate neural network parameters into interpretable causal models, science risks descending into a form of "digital alchemy," where researchers blindly trust predictive engines without understanding the principles driving their output.
To bridge this divide, computer scientists are developing hybrid systems that merge data-driven deep learning with symbolic AI—a domain known as neuro-symbolic AI. By embedding established physical laws, conservation principles, and mathematical symmetries directly into the architecture of neural networks, neuro-symbolic systems force algorithms to operate within the bounds of known physical laws. These constrained models not only require significantly less training data, but they also produce outputs that are inherently interpretable and physically consistent. Ultimately, the future of AI-driven scientific discovery may not lie in replacing human scientific reasoning, but in expanding it through hybrid systems that pair machine pattern recognition with human conceptual synthesis.
Questions
Questions 1–7
Do the following statements agree with the information given in Reading Passage 1?
In boxes 1–7 on your answer sheet, write:
TRUE if the statement agrees with the information
FALSE if the statement contradicts the information
NOT GIVEN if there is no information on this
Questions 8–13
Complete the summary below.
Choose NO MORE THAN TWO WORDS from the passage for each answer.
Engineering Innovations in Offshore Mariculture
Offshore mariculture overcomes environmental degradation by placing fish farms in high-energy marine environments. To survive harsh weather, modern operations use submersible cages made with frameworks of and copper-alloy mesh. When satellite monitoring indicates approaching storms, these structures flood their to lower the cages beneath dangerous wave motion. These remote installations achieve energy independence by utilizing floating buoys powered by , wind, and wave energy. Data regarding water conditions and fish mass is sent to shore via . To reduce financial losses from unused food, algorithms operate to distribute pellets efficiently. However, broad adoption is still hindered by high along with complex jurisdictional issues in international waters.
Questions 14–19
Reading Passage 2 has six paragraphs, labeled A–F.
Which paragraph contains the following information?
Write the correct letter, A–F, in boxes 14–19 on your answer sheet.
Questions 20–22
Choose THREE letters, A–F.
Write the correct letters in boxes 20–22 on your answer sheet.
Which THREE astronomical phenomena or calendar calculations were tracked using the dials and gears of the Antikythera Mechanism?
Questions 23–26
Complete the sentences below.
Choose NO MORE THAN TWO WORDS from the passage for each answer.
Questions 27–31
Choose the correct letter, A, B, C, or D.
Write the correct letter in boxes 27–31 on your answer sheet.
Questions 32–36
Do the following statements agree with the claims of the writer in Reading Passage 3?
Write YES, NO, or NOT GIVEN.
Questions 37–40
Complete the summary using the list of words, A–H, below.
Write the correct letter, A–H, in boxes 37–40 on your answer sheet.
The role of artificial intelligence in research has sparked an epistemological debate. Traditional science seeks mechanistic explanation, whereas deep learning relies on hidden parameters generating predictions through . Pragmatic instrumentalists prioritize , arguing that complex natural patterns need not conform to human cognitive limits. However, realists warn that ignoring causal pathways leads to based on unintended dataset artifacts. To resolve this, researchers are developing , which integrate fundamental physical principles directly into AI architectures to ensure interpretable, reliable results.