Fae Amiro is a PhD candidate at McMaster University in Hamilton, Ontario. Her research focus is Roman portraiture, and she is currently writing a dissertation on the portraiture of the Imperial house during the reign of the emperor Hadrian, which addresses broader questions of portrait type creation and the dissemination of sculpture throughout the Roman empire. She was a participant in the 2017 Eric P. Newman Graduate Seminar.
The empress Sabina is not a figure who is frequently given much attention, due in part to her lack of prominence in the literary record. She was wife of the emperor Hadrian and they are said to have had an unhappy marriage, but not much else is known. Her coinage, however, has received more scholarly attention, because it was issued in larger numbers than that of any previous empress and features a good deal of variety in its portraiture. The question of the true chronology of her coinage has been debated for ninety years. However, few have addressed the reasons behind the changes observable in the coinage, in particular the impetus behind its start date and the introductions of new types.
In order to address these problems, I conducted a die study of the aurei which display the portrait of the empress Sabina. This had not previously been done and is the best way to form a relative chronological sequence for coinage. The die-link sequence confirms the following chronology for the portrait types which appear on the aurei. First is a type called the turban, dating to 128 C.E. (Fig. 1). The next type is Sabina’s main portrait type, the queue, which was probably introduced in the year 131 C. E. (Fig. 3). The Aphrodite type comes next, around 133/134, and was in use until her death and shortly thereafter. Following her death in 136/7, she was consecrated as a diva and a posthumous issue was created to commemorate this.
So, this answers the question of the true sequence of the types. The reasons for the creation of the last two types, the Aphrodite and posthumous types, are well understood. The Aphrodite is represented in a classicizing style, which is associated with Hadrian’s return to Rome after his trips in the east. The posthumous type was created to commemorate her consecration.
The impetus behind the creation of the other types is harder to address. The motivation behind the start of coining for the empress in 128, eleven years into Hadrian’s reign, is unclear. Previously scholars believed that it was because Sabina gained the title of Augusta in that year, but this has been proven incorrect by Eck and the results of the die study. Most likely a number of factors came together at the right time to inspire this change: the ten year anniversary of a reign was a common time for coinage reform, the imperial couple had just returned from a trip abroad and were about to embark on another one, there were no other Augustae alive at the time, and Sabina’s presence on coinage may have helped advertise the family’s prestige, given her relation to the imperial family of the previous dynasty. This last point is supported by the style of the portrait, which resembles that of her mother, Matidia, who was Trajan’s niece (Fig. 2).
Scholars have previously believed that the queue type was introduced to form a visual connection between Sabina and her predecessor, Plotina. However, there are a number of problems with this assessment. This message would have been redundant, since the turban already showed dynastic continuity, and untimely, since Plotina had died eight years previous. A side by side comparison shows that the very assertion that they look alike has been overstated, especially given the prevalence of ponytail-style hairdos among women at this time (Fig. 4). The motivation for the creation of the type is more likely the opposite, that it actually represents a stylistic departure from the previous dynasty and the introduction of a uniquely Hadrianic style.
More work needs to be done, but the results so far show that Sabina’s life events, particularly in association with Hadrian’s imperial travels, had an effect on the appearance of her coin portraits.
Gregory Callaghan is a PhD Candidate in Ancient History at the University of Pennsylvania, and a Kolb Fellow of the University of Pennsylvania Museum of Archaeology and Anthropology. Greg has a wide range of research interests, including Roman Republican politics and the evolution of Athenian democratic institutions. His main focus, however, is the international relations of the Hellenistic Period. His dissertation applies IR theories of status hierarchy and regional powers to the growth of the Attalid kingdom, examining Attalid foreign benefactions through epigraphy and archaeology.
One of the traditional lectures of the Eric P. Newman Graduate Seminar in Numismatics is entitled “Cistophoric Mysteries.” It is an apt name for a lecture dedicated to the cistophoric coinage of Asia Minor. Known as one of the ugliest coins ever produced, the cistophorus was introduced by the Attalid dynasty on a reduced silver standard, somewhere between 190 and 160 BCE. And that is more or less all that we agree on, with the exact date of its introduction, its nature of production, and to what degree it was a closed currency system still debated. But if we still scratch our heads at the nature of the early cistophoric coinage, this is nothing compared to our ignorance of the late cistophoric coinage. From ca. 105–60 BCE, a generation after the end of the Attalid dynasty, several Anatolian mints issued a second series of cistophoric coinage, which was followed by issues produced by these same mints under the direction of Roman magistrates. It is only now that these later issues of cistophori have begun to receive the scholarly attention they deserve. Helped in particular by the generous bequeathal of Rick Witschonke’s collection of cistophori to the ANS—undoubtedly the finest single collection of the coins in the world—Lucia Carbone and William Metcalf have forthcoming studies that investigate the late cistophori of Tralles and the Proconsular Cistophori, respectively.
My project for the graduate seminar picks up on these recent studies by examining the late cistophori of Laodiceia-ad-Lycum. This city is described by Strabo as one of the two great cities of Phrygia, along with Apameia—another late cistophoric mint. Hoard evidence suggests that it began minting cistophori around 90 BCE, and continued into the 60s BCE, when the mint switched from seemingly civic issues to issues stamped with the authority of the Roman proconsuls. My die study of the coinage shows that the mint had a stable output of cistophori through both periods of production. This may seem unremarkable, until we place this in a greater context, both chronologically and geographically. Under the Attalids, Laodiceia produced only 4 identified obverse dies, less than 2% of total cistophoric production. Yet it produced 46 identified obverse dies in the late cistophoric period. Furthermore, in the period of late cistophoric production, Laodiceia produced less than 10% of the total cistophori minted in the province, but it produced over 20% of the proconsular cistophori.
These substantial changes between periods of production need be explained. Until we have more complete studies of all the mints of the late cistophoric period, we can do little more than speculate. But it seems that around 90 BCE, something occurred at the local level that resulted in Laodiceia springing from a mint hardly worth mentioning, to one producing at the level of its more established neighbor, Apameia. In the decades that followed, as we become more certain of Roman oversight, it is clear that a Roman provincial policy developed that maintained the cistophoric production of Laodiceia, while dramatically cutting the production of other mints. My study of Laodiceia can do little more than draw attention to the existence of these situations, and cannot pinpoint the explanation behind them, but it does provide an important piece of the puzzle that I hope can be completed with the addition of further pieces as more of the late cistophoric mints are subjected to similar scrutiny.
The dating, function, and iconography of Roman lead tokens from Spain have been objects of speculation among scholars for decades. Several of these tokens, with weights ranging from 4–400 grams, have been found in the Spanish region of Cordova, once part of the Hispania Baetica, an area known in Roman times for silver mines. Spanish silver mines were one of the most important sources of silver bullion for Rome, and the connected smelting activities took place on such a huge scale that the lead pollution generated by them is still traceable in the Greenland ice core. At the same time, Baetica was also an important producer of olive oil, traded all over the Mediterranean Sea. Spanish lead tokens then, made out of a by-product of silver smelting but possibly also connected to agriculture, represent a useful yet poorly understood tool to understand the economic organization of this province.
The Richard B. Witschonke Collection at the ANS includes 16 specimens of these tokens, nine of which remain unpublished. One of them (fig. 1) is a unique piece, part of lot of 10 Spanish lead tokens offered for sale in CNG MBS 67 on September 22, 2004 (lot nos. 1070–1079). The CNG catalogue offers the following description:
Obv. Nude male walking left, carrying bell(?) and shovel over his shoulder; P · S across field; all within wreath. Rev. Harrow (or miner’s axe?). Weight: 166.78 g
The identity of the man represented on the obverse, together with the function of the objects he is carrying, is a mystery. Is he a miner, carrying a shovel? This is the interpretation offered by F. Casariego, G. Cores, and F. Pliego, who first published this piece in their catalogue of Iberian lead tokens from Roman times. They classified this piece as part of the series de las minas (“mines series”), conventionally related to the Roman mining operations in Baetica. These mine tokens (figs. 2, 3, 4) are usually characterized by the presence of a man with a “shovel” (a conventional term; it is unclear what this is).
This representation closely resembles the miners portrayed on the Linares bas-relief (fig. 5). Moreover, some tokens of the series de las minas were found in the Roman mines of El Maderero (fig. 6) and of Posadas (fig. 7), both in the Baetican district of Cordova. The archaeological context suggests a dating in the first century BC for these tokens. According to this interpretation, these tokens may have served as a ‘company coinage’ for these mines, a practice well attested in modern times. This token and the others of the “mines series” would therefore be one of the first instances of this use of tokens.
However, some elements in the iconography of the token represented in fig. 1 do not seem to match this interpretation. The bell carried by the man with the “shovel” and the arrow on the reverse need to find an explanation. A possible solution for this enigma does not come from Spain, but from Central Italy.
In a series of articles, C. Stannard showed the certain iconographical relationship between lead tokens from Baetica and local bronzes from Central Italy. The motif of the man with the “shovel” is attested in the area of Minturnae, Naples, and Pompeii, where no connection to mining activities can be made (fig. 8). The man with the “shovel” was probably not a miner, after all. As represented in fig. 4, the most frequent iconography of this figure is a walking man, either naked or wearing a short tunic, carrying the “shovel.” In the Italian material, he often also carries an askos, an oil or wine jar; in the Baetican, a bell (as in the case of the token in fig. 1). Could the man with the “shovel” be a farmer? The farming context could help explaining the presence of a harrow on the reverse of our token. Moreover, M. P. García-Bellido argues that the letters P · S, appearing on the token at ANS and on other ones of the same series, could be interpreted as P(ublica) S(ocietas), a State-owned enterprise exploiting oil-production in Baetica. According to this second interpretation then, the tokens of the series de las minas were used as a “company coinage” in an agricultural context, not in a mining one.
The iconographical similarities between Baetican tokens and Italian bronzes bear testimony to the active commercial relationships between Italy and Baetica in the Age of High Empire (first–second centuries AD), especially wine and oil trade. Mount Testaccio in Rome, an artificial mound composed almost entirely of testae, fragments of broken oil and wine amphorae dating from the first– third centuries AD (figs. 10, 11) bears testimony to the enormous scale of this trade. While researching Mount Testaccio’s amphora stamps, B. Mora Serrano (fig. 9) noticed the correspondence between the names appearing on some tokens of the series de las minas and the ones on amphora stamps from Testaccio. He therefore argued that at least some of the Iberian lead tokens of the series de las minas are connected to the transport of the Spanish olive oil to Rome. It follows that the man with the “shovel” on the unique piece of the Richard Witschonke Collection would not be an Iberian miner, but rather an Italo-Baetican farmer, probably occupied in producing wine and oil to export to Italy.
However, neither the presence of a bell nor the generously ithyphallic representations (cf. fig. 4) of the man with the “shovel” are addressed by this interpretation. C. Stannard argues that these elements could be explained if these figures were mimes. The Roman mime differed from Greek Comedy in that actors did not wear masks, as in the images on the Iberian lead. According to Stannard’s hypothesis, the man with the “shovel” represents a mime, a decorative element on tokens that were used as a “company coinage” in the context of an Italo-Baetican oil-trade enterprise.
In sum, the identity of the man with the “shovel” on the token presented in fig. 1 raises historical and iconographic questions that show the strength of the commercial and cultural interconnections within the Roman world. Were the tokens of the series de las minas really connected to mining activities, as their findspots seem to suggest? Or were they connected to the trade of Spanish oil, as B. Mora Serrano posits? The debate is still open.
Two elements still need further interpretation:
Even if not univocally linked to mines, some tokens of the series de las minas did circulate in mining areas. It is therefore not possible to entirely dismiss the “mining” interpretation.
The findspots of some tokens of the series de las minas show that these kind of tokens were already circulating during the first century BC, so they could not be directly linked to the Spanish oil trade of the first and second centuries AD.
Not all the questions are solved, then. The mystery of “our” man with the “shovel” is still intact.
New hypotheses on the iconography and the function of the man with the “shovel” and the function of the fascinating Spanish lead tokens will be formulated at the interdisciplinary conference “Tokens: Culture, Connections, Communities” at Warwick University (June 8–10, 2017), where all the published and unpublished lead tokens from the Richard W. Collection will be presented.
The ANS is pleased to announce a new interface for Online Coins of the Roman Empire (OCRE), which allows non-specialists, hobbyists, collectors, archaeologists, and others to browse Roman Imperial coins by image for free online. People can compare the coins in their collections or those coins recovered from archaeological excavations against diagnostic specimens in OCRE. The OCRE project received $300,000 from the National Endowment for the Humanities (NEH) in 2014.
The “Identify a Coin” interface works on any device, from computers to tablets to smart phones. Users can begin browsing Roman Imperial coinage right away, or can filter search criteria by portrait, material, and even legends, which includes the ability to enter all or part of a legend as well as marking illegible characters. The portraits are listed chronologically, first by dynasty, and then by personage within the dynasty (including empresses and children). In many cases, examples of portrait images are available in gold, silver, and bronze varieties, as well as worn examples that one may encounter with stray finds or excavation. More than one material may be chosen, which is useful for later Roman coinage, when severe wear makes it difficult to distinguish between what Roman Imperial Coinage has designated as “silver,” “bronze,” or “billon.” By clicking the left and right arrows below the image, it is possible to scroll through available portraits, which may show several phases of portraiture, such as Nero, who grew from a teenager into adulthood over the course of his reign.
OCRE’s “Identify a Coin” tool offers an easy way in to one of the most complete depictions of numismatic Imperial portraiture online, and the ANS hopes that it will also prove itself a useful art historical tool to trace the development of Roman portraiture from the Augustan period through the Soldier Emperors to the Tetrarchy until the end of the Roman Empire.
Ethan Gruber, the ANS’s Director of Data Science, created the interface, and ANS Curatorial Assistant Disnarda Pinilla identified all of the portraits used in the tool. The ANS’s Associate Curator, Gilles Bransbourg, has overseen the OCRE project from its inception. Although primarily drawn from the ANS’s permanent collection of Roman Imperial coins, other specimens are included from the Münzkabinetts in both Berlin and Vienna, as well as from the Fralin Museum at the University of Virginia.
Last month, on November 7, 2016, the internationally acclaimed Canadian artist Leonard Cohen died of cancer. He was regarded as a man of many talents, who painted, wrote novels, poetry, and the songs for which he was best known. He was a man of art, culture, and ideas, who appreciated the value of both Manhattan and Berlin.
Within the large oeuvre left behind by Cohen, the song Hallelujah stands out for its great popularity and the life of its own that it has taken on in the hands of the many other musicians (now more than 300 in multiple languages according to Wikipedia) who have played and modified its lyrics since it was first released in 1984. However, all versions begin with Cohen’s original lament “Now I’ve heard there was a secret chord / That David played, and it pleased the Lord / But you don’t really care for music, do you?” and this has prompted the topic for this edition of the ANS blog. Regardless of whether a numismatist does care for music (Cohen’s or anyone else’s) or not, David’s secret chord and those of other lyric poets have made an impact on coins from antiquity up to modern times.
David is depicted playing his chord on the enigmatic Irish St. Patrick coinage of the seventeenth century, some of which was carried off to New Jersey to serve as halfpence in the cash-starved colony. Its production in two denominations (or one that was later reduced in weight?), date of issue, meaning of its iconography, and the circumstances of its arrival in New Jersey have captured the imagination of Colonial American numismatists for decades and even became the topic for an ANS Coinage of the Americas Conference in 2006. The famous king also plays on contemporary coins of Nuremberg and the Papal States, as well as on psalmenpfennige (medallic awards for the completion of Protestant religious education, which included the memorization of Psalms—the Biblical songs attributed to David).
Harps appear without players on English silver coins struck for use in Ireland in the sixteenth century, but it is not clear whether any reference to King David was intended in this heraldic emblem or whether the instrument alluded only to contemporary Irish culture, which held its native harpers in high esteem. David’s harp (indicated by the winged female column symbolizing the unearthly beauty of its music), occurs on English halfpence produced for Ireland English halfpence produced for Ireland in the eighteenth century—perhaps not coincidentally after more than 100 years of repressive policies had all but crushed the native tradition of harping in Ireland. A pointedly Celtic harp (Irish cláirseach) has been used on all Irish coins, including the current euro, since the creation of the Irish Free State (Republic of Ireland after 1948) in 1922.
Although it is regularly described and depicted as a harp in medieval and modern texts and artworks, David’s stringed instrument was actually a form of lyre known in Hebrew as the kinnor. A related instrument, the nebel was regularly played as part of celebratory worship in the Jerusalem Temple. The connection of these instruments to the Temple and to David lies behind their prominent depiction on coins struck by Jewish rebels against Rome during the disastrous Bar Kokhba War (AD 132-136). This bloody conflict erupted when the emperor Hadrian sought to refound Jerusalem (already destroyed by Titus in AD 70) as a pagan city. The kinnor and nebel of the Bar Kokhba coins had a dual purpose. They evoked the longing memory of days when the Temple still stood and great Jewish kings ruled the land while casting Simon bar Kokhba, the leader of the revolt, as a Messianic figure who might lead his people to victory and restore the Temple.
The chelys (Latin testudo) and kithara of the Greeks and Romans appear to have been the rough equivalents of the kinnor and nebel, respectively. The former, which included a sound box made from a tortoise shell or wood formed into the shape of a shell, was said to have been discovered by the god Hermes. While traveling along a riverbank, he was attracted by a beautiful sound and when he went to investigate he found that the wind was blowing tendons that had been stretched across a tortoise shell. From this he fashioned the first chelys to be played by gods and men. The kithara, however, was a more elevated instrument of wooden construction associated with Apollo and the Muses as patrons of culture and the arts. Indeed, music lessons on the kithara or chelys were a staple of state education programs for citizens of the ancient Greek cities. The important role of music in Greek education is underlined by coins of the Bithynian king Prusias II that depict Chiron, the centaur tutor of Herakles, playing a kithara. The instrument is the only element of the type that allows the viewer to identify the subject as the educator Chiron and not some other centaur.
Apollo is a ubiquitous deity on Greek coins struck in the Classical and Hellenistic periods and even under the Roman Empire, who often appears in his role as Kitharoidos—the kithara player. His kithara is also depicted on coins, often paired with the head of its divine player or by itself, as on early coins of Delos, the island of Apollo’s birth. Less commonly, even great human lyric poets or the Muses who inspired them to greatness appear on Greek and Roman coins. Sappho is shown playing a kithara on coins of Mytilene in the Roman period as a means of advertising the cultural importance of the city while Terpsichore, Kalliope, and Erato, the respective Muses of choral song, epic poetry, and love poetry appear holding a kithara (Kalliope) or chelys (Terpsichore and Kalliope) on coins of the Roman Republican moneyer Q. Pomponius Musa as an extended pun on his cognomen Musa.
While Leonard Cohen attributes a single secret chord to David in his song, the numerous symbolic uses to which harps, lyres, and their players were put on coins of ancient and more modern times would seem to suggest that through the ages there were in fact many such chords aimed at pleasing the mortal as well as the divine.
Today’s post is written by Jeremy Simmons, a PhD candidate in the Classical Studies program at Columbia University. He has written on the topic of Indo-Roman trade, and in particular, the spice trade in antiquity. His dissertation will look specifically at patterns of consumption in the larger Indian Ocean trade network, including the engagement between Indian monetary systems and imported Roman coinage. His project, as a participant of the the 2016 Eric P. Newman Graduate Summer Seminar, focuses on Western Kṣatrapa coins.
For my ANS Seminar Project, I decided to look at silver coins of the Western Kṣatrapas, who ruled in the modern day Indian states of Gujarat and Madhya Pradesh from the mid-first to early fifth centuries CE (Figure 1). These coins have been indispensable for reconstructing the chronology of the Western Kṣatrapa kings, as well as the line of succession, due to the presence of dates (in the Śaka Era) and patronymics on coins. However, the feature of these coins that drew me to this project is the obverse legend, which appears to be written in a script that mixes Greek and Roman characters at random.
These legends have been little discussed in scholarship (as opposed to the Brāhmī legends on the reverse), and have been variously labeled “Greek,” “corrupt/pseudo/blundered Greek,” or “Greco-Roman,” without much consideration of the larger implications of these different descriptions. In fact, it is a general practice to not record the obverse legends of Kṣatrapa coins in catalogues or other publications: cataloguers justify their actions by stating that the legends become corrupt over time and cease to have any meaning; and those publishing or auctioning a single coin tend not to transcribe the obverse legend at all.
This lack of scholarly attention arises from the assumption that the obverse legends at one point communicated written language—specifically, the coins of the early kings like Nahapāna, which have Greco-Roman script transliterations of the Prakrit reverse legends (Figure 2)—but that later die-cutters, due to their lack of skill or knowledge of Greek and Latin, merely rendered corrupt versions as a form of ornamentation. I believe this narrative of decline, first suggested over a century ago, is not only based on the limited evidence of early collections, but has been perpetuated by a regrettable practice of not recording positive data.
In order to correct the treatment of obverse legends on Kṣatrapa coins in scholarship, I have created a database of Greco-Roman obverse legends found on silver coins minted during the three-and-a-half centuries of Kṣatrapa rule (from Nahapāna to Rudrasiṃha III). I gathered these legends from specimens presented in various museum catalogues, auction listings, and publications. While this task involved some difficulties due to the damaged nature of most obverse legends (Figure 3) and the poor quality of photographs, I managed to assemble a large corpus of data in order to supplement existing descriptions and serve as the basis of my initial observations.
It is my hope that these observations will contribute to the following aims: 1) determining which paleographic features of the obverse script are Greek versus Roman, in order to mitigate the problem of variable terminology; 2) outlining possible sources of inspiration for these legends, whether it be local precedents (e.g., Indo-Greek, Indo-Scythian, and Indo-Parthian), imported Roman coinage, or imitation Roman coinage produced in India (Figures 4 and 5); 3) uncovering any evidence of conscious design behind these legends (as opposed to mindless copying as previously suggested), indicated by similar patterns of letters, standardization of legends, etc.; 4) and, most importantly, speculating why the Kṣatrapa kings would design coins with unreadable obverse legends alongside very legible Brāhmī legends and numerals.
At the very least, it is my hope to show the merits in investigating elements of a coinage tradition that many have disregarded as “meaningless.”
[Today’s post is authored by Sam Caldis (Brown University) who is taking part in the 2016 Eric P. Newman Graduate Seminar in Numismatics. Sam’s proposed doctoral dissertation topic is collegial rule and imperial power-sharing in the Roman Empire during the third and fourth centuries. Aside from his seminar project, Sam is spending time at the ANS researching the coins and public image programs of fourth-century Roman emperors, studying for his preliminary exams.]
Coins played a significant role in shaping the public image of the Roman imperial family. The majority of the coins which left the central imperial mints featured the emperor himself and emphasized his activities and virtues, from (re)conquering Britain to his close friendship with Hercules. However, emperors did not always appear solo on their coinage- wives, ancestors, and descendants could all be placed on coins to highlight other marks of distinction for the regime, such as proclaiming ancestry or demonstrating the security of the succession. Although there has recently been a great deal of interest in the place of members of the imperial house on coinage, one aspect which has been largely ignored is the development of familial types – coins which display at least two members of the imperial family together on either the obverse or reverse of a coin.
These types appear early on – Augustus minted coins with his chosen successors, Gaius and Lucius Caesar, together on the reverse. By the end of the reign of Septimius Severus (193-211), familial types had become more common. This is in large part due to the prominence of his two potential heirs, Caracalla (198-211) and Geta (209-211), and their powerful mother, Julia Domna.
The members of this imperial household were mixed and matched to produce a variety of familial types. The most common familial types depict Severus’s sons on the reverse, usually with a legend emphasizing concordia, or harmony. This may have been intended to reinforce the dynasty’s stability to the people of the empire, though it may also have been an unsubtle message from the imperial court to a pair of brothers who were notoriously always at odds.
The rapid turnover of emperors in the third century led to further experimentation with familial types. For example, Gallienus (253-268) has no familial types with his sons and co-emperors Valerian II (253-257) and Saloninus (258-260), but the latter are included in familial types on their mother Salonina’s coins. Instead of appearing as teenage Caesars, Valerian II and Saloninus are depicted with a third child quite literally no bigger than their mother’s knee, emphasizing the fertility and youth of the dynasty.
Perhaps the most interesting shift in familial types during this period is the movement of the family portrait from its usual place on the reverse to the obverse, in some cases. Before the third century, only Augusts and Nero ever produced coinage with an obverse familial type, and these were exceedingly rare. By the end of the third century, the obverse became the preferred location for the family portrait on familial types. Furthermore, the emperor’s solo portrait on the reverse came to be replaced by more traditional imagery, such as deities. A significant number of familial types now looked like “normal” coins with the single emperor on the obverse being replaced by a corporate image of two members of the imperial house. In the rapidly changing political world of the third century CE, emperors experimented with the presentation of their public image in a variety of media as they fought to maintain their position. Familial coin types were one of part of their toolkit, one which received significant attention and innovation.